2010-07-26 Martin Robinson <mrobinson@igalia.com>
[WebKit-https.git] / WebCore / ChangeLog
1 2010-07-26  Martin Robinson  <mrobinson@igalia.com>
2
3         Reviewed by Xan Lopez.
4
5         [GTK] Get rid of libgdom
6         https://bugs.webkit.org/show_bug.cgi?id=42378
7
8         Remove all build steps for building GObject DOM bindings.
9
10         * GNUmakefile.am: Remove GObject DOM bindings build.
11
12 2010-07-26  Mark Rowe  <mrowe@apple.com>
13
14         Reviewed by Dan Bernstein.
15
16         <http://webkit.org/b/42990> REGRESSION (r63854): Safari RSS layout is broken due to changes in vendor prefix handling
17
18         In r63854 support for properties with the -khtml vendor prefix was removed. This breaks the layout of
19         Safari's RSS interface, several Dashboard widgets that are included with Mac OS X, and at least one
20         popular third-party Mac OS X application. This change will need to be revisited in a manner that has
21         a clearer backwards-compatibility strategy. <http://webkit.org/b/42093> will be reopened to track that.
22
23         * css/CSSParser.cpp:
24         (WebCore::cssPropertyID): Revert the WebCore change from r63854.
25
26 2010-07-26  Brady Eidson  <beidson@apple.com>
27
28         Reviewed by Sam Weinig.
29
30         Add ability for WK2 to register a scheme to always be displayed as an empty document
31         https://bugs.webkit.org/show_bug.cgi?id=42995
32
33         No new tests. (Currently not possible to test in WebKit1 DRT, can be testing in future WK2 tester)
34
35         * WebCore.exp.in:
36         
37         * loader/MainResourceLoader.cpp:
38         (WebCore::shouldLoadAsEmptyDocument):
39         
40         * platform/SchemeRegistry.cpp:
41         (WebCore::emptyDocumentSchemes):
42         (WebCore::SchemeRegistry::registerURLSchemeAsEmptyDocument):
43         (WebCore::SchemeRegistry::shouldLoadURLSchemeAsEmptyDocument):
44         * platform/SchemeRegistry.h:
45
46 2010-07-26  Patrick Gansterer  <paroga@paroga.com>
47
48         Reviewed by Kent Tamura.
49
50         [WINCE] Port ClipboardUtilities to WinCE
51         https://bugs.webkit.org/show_bug.cgi?id=42929
52
53         Make ClipboardUtilitiesWin more portable and remove the global namespace prefix
54         from GlobalLock/GlobalUnlock, since they are only macros on WinCE.
55
56         * platform/win/ClipboardUtilitiesWin.cpp:
57         (WebCore::urlFromPath):
58         (WebCore::getWebLocData):
59         (WebCore::createGlobalData):
60         (WebCore::getURL):
61         (WebCore::getPlainText):
62         (WebCore::fragmentFromHTML):
63
64 2010-07-26  Patrick Gansterer  <paroga@paroga.com>
65
66         Reviewed by Adam Roben.
67
68         [WINCE] Buildfix for BitmapImage.h
69         https://bugs.webkit.org/show_bug.cgi?id=42931
70
71         Use AffineTransform insted of TransformationMatrix.
72
73         * platform/graphics/BitmapImage.h:
74
75 2010-07-26  Dirk Schulze  <krit@webkit.org>
76
77         Unreviewed sort of XCode project file.
78
79         * WebCore.xcodeproj/project.pbxproj:
80
81 2010-07-26  David Kilzer  <ddkilzer@apple.com>
82
83         <http://webkit.org/b/42982> Streamline Inspector Source build phase script should use CONFIGURATION not BUILD_STYLE
84         <rdar://problem/6341764>
85
86         Reviewed by Mark Rowe.
87
88         * WebCore.xcodeproj/project.pbxproj: Switched Streamline
89         Inspector Source build phase script to use CONFIGURATION instead
90         of BUILD_STYLE.
91
92 2010-07-23  Ojan Vafai  <ojan@chromium.org>
93
94         Reviewed by Darin Adler.
95
96         crash in insertParagraph in a contentEditable list
97         https://bugs.webkit.org/show_bug.cgi?id=42919
98
99         Don't try to break out of a list item if the list it's in
100         is the rootEditableElement.
101
102         * editing/CompositeEditCommand.cpp:
103         (WebCore::CompositeEditCommand::breakOutOfEmptyListItem):
104
105 2010-07-26  Marcus Bulach  <bulach@chromium.org>
106
107         Reviewed by Jeremy Orlow.
108
109         Moves IDBKeyRange to use IDBKey rather than SerializedScriptValue / IDBAny.
110         https://bugs.webkit.org/show_bug.cgi?id=42583
111
112         Tests: existing storage/indexeddb/idb-keyrange.html must pass.
113
114         * storage/IDBKeyRange.cpp:
115         (WebCore::IDBKeyRange::IDBKeyRange):
116         * storage/IDBKeyRange.h:
117         (WebCore::IDBKeyRange::create):
118         (WebCore::IDBKeyRange::left):
119         (WebCore::IDBKeyRange::right):
120         * storage/IDBKeyRange.idl:
121         * storage/IndexedDatabaseRequest.cpp:
122         (WebCore::IndexedDatabaseRequest::makeSingleKeyRange):
123         (WebCore::IndexedDatabaseRequest::makeLeftBoundKeyRange):
124         (WebCore::IndexedDatabaseRequest::makeRightBoundKeyRange):
125         (WebCore::IndexedDatabaseRequest::makeBoundKeyRange):
126         * storage/IndexedDatabaseRequest.h:
127         * storage/IndexedDatabaseRequest.idl:
128
129 2010-07-26  Simon Fraser  <simon.fraser@apple.com>
130
131         Reviewed by Anders Carlsson.
132
133         Composited layers don't scroll in WebKit2
134         https://bugs.webkit.org/show_bug.cgi?id=42771
135
136         Rename two methods on RenderLayerCompositor to make their use more clear, and call them
137         when the FrameView gets resized and scrolled.
138
139         * page/FrameView.h:
140         * page/FrameView.cpp:
141         (WebCore::FrameView::setFrameRect): override so we know when the size of the FrameView changes, so that the
142         RenderLayerCompositor can update its clipping layer.
143         (WebCore::FrameView::scrollPositionChanged): Call compositor()->frameViewDidScroll().
144
145         * rendering/RenderLayerBacking.cpp:
146         (WebCore::RenderLayerBacking::updateAfterWidgetResize): updateContentLayerOffset() was renamed.
147
148         * rendering/RenderLayerCompositor.h:
149         * rendering/RenderLayerCompositor.cpp:
150         (WebCore::RenderLayerCompositor::frameViewDidChangeSize): Renamed from updateContentLayerOffset().
151         (WebCore::RenderLayerCompositor::frameViewDidScroll): Renamed from updateContentLayerScrollPosition().
152         (WebCore::RenderLayerCompositor::shouldPropagateCompositingToEnclosingIFrame): Propagate compositing on Mac
153         for any FrameView that doesn't have a native widget, which is the case with WebKit2.
154         (WebCore::RenderLayerCompositor::requiresScrollLayer): New method to tell us whether we need to handle
155         scrolling (and clipping) ourselves.
156         (WebCore::RenderLayerCompositor::ensureRootPlatformLayer): Do geometry flipping on the root platform
157         layer only if we're not doing our own scrolling, and use the new requiresScrollLayer() method.
158         (WebCore::RenderLayerCompositor::attachRootPlatformLayer): Send out the rootPlatformLayer(),
159         so that it returns the clipping layer if we have one.
160
161 2010-07-26  Tony Gentilcore  <tonyg@chromium.org>
162
163         Reviewed by Darin Fisher.
164
165         Move DocumentLoadTiming struct to a new file
166         https://bugs.webkit.org/show_bug.cgi?id=42917
167
168         Also makes DocumentLoadTiming Noncopyable.
169
170         No new tests because no new functionality.
171
172         * GNUmakefile.am:
173         * WebCore.gypi:
174         * WebCore.vcproj/WebCore.vcproj:
175         * WebCore.xcodeproj/project.pbxproj:
176         * loader/DocumentLoadTiming.h: Added.
177         (WebCore::DocumentLoadTiming::DocumentLoadTiming):
178         * loader/DocumentLoader.h:
179         * loader/FrameLoader.cpp:
180         * loader/FrameLoaderTypes.h:
181         * loader/MainResourceLoader.cpp:
182         * page/Timing.cpp:
183
184 2010-07-26  Mark Rowe  <mrowe@apple.com>
185
186         Fix an #if in WebCore.exp.in to correctly detect whether SVG filters are enabled.
187
188         * WebCore.exp.in:
189
190 2010-07-26  Steve Block  <steveblock@google.com>
191
192         Reviewed by Jeremy Orlow.
193
194         DeviceOrientation event listeners should never be called synchronously from addEventListener()
195         https://bugs.webkit.org/show_bug.cgi?id=42304
196
197         No new tests as there are currently no implementations to test. Tests will be
198         added once LayoutTestController methods for providing a mock implementation are
199         in place, see Bug 39589.
200
201         * dom/DeviceOrientationController.cpp:
202         (WebCore::DeviceOrientationController::DeviceOrientationController):
203         (WebCore::DeviceOrientationController::timerFired):
204         (WebCore::DeviceOrientationController::addListener):
205         (WebCore::DeviceOrientationController::removeListener):
206         (WebCore::DeviceOrientationController::removeAllListeners):
207         * dom/DeviceOrientationController.h:
208
209 2010-07-26  Andreas Kling  <andreas.kling@nokia.com>
210
211         Reviewed by Tor Arne Vestbø.
212
213         [Qt] Canvas: Optimize GraphicsContext::addPath()
214
215         Avoid "deep concatenating" the canvas path onto the GC path
216         if the GC path has no elements.
217
218         * platform/graphics/qt/GraphicsContextQt.cpp:
219         (WebCore::GraphicsContext::addPath):
220
221 2010-07-23  Stephen White  <senorblanco@chromium.org>
222
223         Reviewed by Darin Fisher.
224
225         Implement OpenGLES2 helper classes.
226         https://bugs.webkit.org/show_bug.cgi?id=42905
227
228         GLES2Canvas and GLES2Texture are some simple OpenGLES2 helper classes
229         for managing state, samplers and textures.
230
231         When this is hooked up, it will be covered by many layout tests.
232
233         * WebCore.gypi:
234         Add new files to chromium build (contents are protected by #if's for now).
235         * platform/graphics/chromium/GLES2Canvas.cpp: Added.
236         (WebCore::affineTo3x3):
237         (WebCore::GLES2Canvas::State::State):
238         (WebCore::GLES2Canvas::GLES2Canvas):
239         (WebCore::GLES2Canvas::~GLES2Canvas):
240         (WebCore::GLES2Canvas::clearRect):
241         (WebCore::GLES2Canvas::fillRect):
242         (WebCore::GLES2Canvas::setFillColor):
243         (WebCore::GLES2Canvas::setAlpha):
244         (WebCore::GLES2Canvas::translate):
245         (WebCore::GLES2Canvas::rotate):
246         (WebCore::GLES2Canvas::scale):
247         (WebCore::GLES2Canvas::concatCTM):
248         (WebCore::GLES2Canvas::save):
249         (WebCore::GLES2Canvas::restore):
250         (WebCore::GLES2Canvas::drawTexturedRect):
251         (WebCore::GLES2Canvas::setCompositeOperation):
252         (WebCore::GLES2Canvas::applyCompositeOperator):
253         (WebCore::GLES2Canvas::getQuadVertices):
254         (WebCore::GLES2Canvas::getQuadIndices):
255         (WebCore::loadShader):
256         (WebCore::GLES2Canvas::getSimpleProgram):
257         (WebCore::GLES2Canvas::getTexProgram):
258         (WebCore::GLES2Canvas::createTexture):
259         (WebCore::GLES2Canvas::getTexture):
260         (WebCore::GLES2Canvas::checkGLError):
261         * platform/graphics/chromium/GLES2Canvas.h: Added.
262         (WebCore::GLES2Canvas::gles2Context):
263         * platform/graphics/chromium/GLES2Texture.cpp: Added.
264         (WebCore::GLES2Texture::GLES2Texture):
265         (WebCore::GLES2Texture::~GLES2Texture):
266         (WebCore::GLES2Texture::create):
267         (WebCore::convertFormat):
268         (WebCore::GLES2Texture::load):
269         (WebCore::GLES2Texture::bind):
270         * platform/graphics/chromium/GLES2Texture.h: Added.
271         (WebCore::GLES2Texture::):
272         (WebCore::GLES2Texture::format):
273         (WebCore::GLES2Texture::width):
274         (WebCore::GLES2Texture::height):
275
276
277 2010-07-26  Jeremy Orlow  <jorlow@chromium.org>
278
279         Reviewed by Steve Block.
280
281         Fix IndexedDB release build on mac
282         https://bugs.webkit.org/show_bug.cgi?id=42377
283
284         * storage/IDBDatabaseImpl.cpp:
285         (WebCore::IDBDatabaseImpl::objectStore):
286
287 2010-07-26  Satish Sampath  <satish@chromium.org>
288
289         Reviewed by Steve Block.
290
291         Add WebKit plumbing to connect speech requests and callbacks between WebCore and the embedder.
292         https://bugs.webkit.org/show_bug.cgi?id=42367
293
294         No new tests, the relevant LayoutTestController bindings will be added in the next patch.
295
296         * page/SpeechInput.cpp: renamed a method.
297         * page/SpeechInput.h: renamed a method.
298         * page/SpeechInputClient.h: added an extra method to optionally let users stop recording once they have spoken.
299         * page/SpeechInputClientListener.h: renamed a method.
300         * page/SpeechInputListener.h: renamed a method.
301
302 2010-07-26  Andreas Kling  <andreas.kling@nokia.com>
303
304         Reviewed by Kenneth Rohde Christiansen.
305
306         [Qt] getImageData(): Single-pass RGB->BGR and un-premultiplication
307         https://bugs.webkit.org/show_bug.cgi?id=42945
308
309         Combine the two operations into a single loop over the pixel data.
310         Yields a considerable FPS gain on http://www.semantix.gr/statue.html
311
312         * platform/graphics/qt/ImageBufferQt.cpp:
313         (WebCore::getImageData):
314
315 2010-07-26  Pavel Feldman  <pfeldman@chromium.org>
316
317         Reviewed by Shinichiro Hamaji.
318
319         Web Inspector: make sure proxy objects are JSON-stringifiable.
320
321         https://bugs.webkit.org/show_bug.cgi?id=42961
322
323         * inspector/front-end/InjectedScript.js:
324         (injectedScriptConstructor):
325
326 2010-07-26  Alexander Pavlov  <apavlov@chromium.org>
327
328         Reviewed by Joseph Pecoraro.
329
330         Inspector should remember resources sorting set by the user
331         https://bugs.webkit.org/show_bug.cgi?id=19208
332
333         Sorting options both for time and size graphs are memorized in a single property,
334         which makes them possible to retrieve in a single message (e.g. useful for remote debugging).
335
336         * inspector/front-end/ResourcesPanel.js:
337         (WebInspector.ResourcesPanel.prototype.populateSidebar):
338         (WebInspector.ResourcesPanel.prototype._settingsLoaded):
339         (WebInspector.ResourcesPanel.prototype._loadSortOptions):
340         (WebInspector.ResourcesPanel.prototype._loadSortOptionForGraph):
341         (WebInspector.ResourcesPanel.prototype._graphSelected):
342         (WebInspector.ResourcesPanel.prototype._changeSortingFunction):
343         (WebInspector.ResourcesPanel.prototype._selectedOptionNameForGraph):
344         (WebInspector.ResourcesPanel.prototype._doChangeSortingFunction):
345         * inspector/front-end/Settings.js:
346         (WebInspector.populateApplicationSettings):
347
348 2010-07-25  Ryosuke Niwa  <rniwa@webkit.org>
349
350         Reviewed by Darin Adler.
351
352         Redo fails after text node is split by SplitTextNodeCommand
353         https://bugs.webkit.org/show_bug.cgi?id=42941
354
355         The bug was caused by SplitTextNodeCommand not implementing doReapply.
356         Implemented doReapply and added insertText1AndTrimText2 to share the code between doApply and doReapply.
357
358         Test: editing/undo/redo-split-text-node.html
359               editing/undo/redo-split-text-node-with-removal.html
360
361         * editing/SplitTextNodeCommand.cpp:
362         (WebCore::SplitTextNodeCommand::doApply): Calls executeApply.
363         (WebCore::SplitTextNodeCommand::doUnapply): Don't delete m_text1, which is later used in doReapply.
364         (WebCore::SplitTextNodeCommand::doReapply): Added.
365         (WebCore::SplitTextNodeCommand::insertText1AndTrimText2): Added.
366         * editing/SplitTextNodeCommand.h:
367
368 2010-07-23  Philippe Normand  <pnormand@igalia.com>
369
370         Reviewed by Eric Carlson.
371
372         [GTK] media/media-can-play-mpeg-audio.html fails
373         https://bugs.webkit.org/show_bug.cgi?id=42895
374
375         Fixed advertizing of mp3, mpeg and mp4 mime-types in the player.
376
377         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
378         (WebCore::mimeTypeCache):
379
380 2010-07-23  Philippe Normand  <pnormand@igalia.com>
381
382         Reviewed by Eric Carlson.
383
384         LayoutTests/media/media-document-audio-size.html failure
385         https://bugs.webkit.org/show_bug.cgi?id=38376
386
387         Advertize audio/x-wav along with audio/wav when appropriate,
388         otherwise WebKit attempts to use a plugin to play the media.
389
390         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
391         (WebCore::mimeTypeCache):
392
393 2010-07-25  Ryosuke Niwa  <rniwa@webkit.org>
394
395         Reviewed by Kent Tamura.
396
397         splitTextAt*IfNeed and splitTextElementAt*IfNeed need to be cleaned up
398         https://bugs.webkit.org/show_bug.cgi?id=42937
399
400         Isolated the code to decide whether or not text node should be split into isValidCaretPositionInTextNode.
401         Moved the condition check out of *IfNeeded methods to applyRelativeFontStyleChange and applyInlineStyle.
402
403         No new tests added since this is a clean up.
404
405         * editing/ApplyStyleCommand.cpp:
406         (WebCore::ApplyStyleCommand::applyRelativeFontStyleChange): Uses isValidCaretPositionInTextNode.
407         (WebCore::ApplyStyleCommand::applyInlineStyle): Uses isValidCaretPositionInTextNode.
408         (WebCore::ApplyStyleCommand::splitTextAtStart): Renamed from splitTextAtStartIfNeeded.
409         (WebCore::ApplyStyleCommand::splitTextAtEnd): Renamed from splitTextAtEndIfNeeded.
410         (WebCore::ApplyStyleCommand::splitTextElementAtStart): Renamed from splitTextElementAtStartIfNeeded.
411         (WebCore::ApplyStyleCommand::splitTextElementAtEnd): Renamed from splitTextElementAtEndIfNeeded.
412         (WebCore::ApplyStyleCommand::isValidCaretPositionInTextNode): Returns true if the position lies within a text node.
413         * editing/ApplyStyleCommand.h:
414
415 2010-07-25  Patrick Gansterer  <paroga@paroga.com>
416
417         Reviewed by Kent Tamura.
418
419         [WINCE] Buildfix for ImageBufferWince
420         https://bugs.webkit.org/show_bug.cgi?id=42928
421
422         * platform/graphics/wince/ImageBufferWince.cpp:
423         (WebCore::):
424         (WebCore::BufferedImage::draw): Pass through ColorSpace parameter.
425         (WebCore::BufferedImage::drawPattern): Ditto.
426         (WebCore::ImageBuffer::platformTransformColorSpace): Added.
427
428 2010-07-25  Patrick Gansterer  <paroga@paroga.com>
429
430         Reviewed by Kent Tamura.
431
432         [WINCE] Buildfix for PopupMenuWin.cpp
433         https://bugs.webkit.org/show_bug.cgi?id=42932
434
435         Add #if !OS(WINCE) around unsupported window messages.
436
437         * platform/win/PopupMenuWin.cpp:
438         (WebCore::PopupMenu::show):
439         (WebCore::PopupMenu::wndProc):
440
441 2010-07-25  Adam Barth  <abarth@webkit.org>
442
443         Reviewed by Eric Seidel.
444
445         Implement adjustMathMLAttributes
446         https://bugs.webkit.org/show_bug.cgi?id=42952
447
448         This is kind of mechanical overkill for adjusting one attribute name,
449         but it's nice to use the same machinery as SVG.
450
451         * html/HTMLTreeBuilder.cpp:
452         * mathml/mathattrs.in:
453
454 2010-07-25  Adam Barth  <abarth@webkit.org>
455
456         Reviewed by Eric Seidel.
457
458         Functions supporting the legacy tree builder are confusing
459         https://bugs.webkit.org/show_bug.cgi?id=42951
460
461         This patch just inlines these functions into their one call site and
462         removes the notImplemented() calls.  We thought we were going to grow
463         these to support the new tree builder, but we took a different path.
464         Previous to this patch, these functions were confusing because they
465         looked like part of the new tree builder.
466
467         * html/HTMLTreeBuilder.cpp:
468         (WebCore::HTMLTreeBuilder::passTokenToLegacyParser):
469         * html/HTMLTreeBuilder.h:
470
471 2010-07-25  Adam Barth  <abarth@webkit.org>
472
473         Reviewed by Eric Seidel.
474
475         Recent HTML5 TreeBuilder Regression?  All canvas tests ASSERT
476         https://bugs.webkit.org/show_bug.cgi?id=42948
477
478         Remove ASSERT and replace with correct branch.  I put this ASSERT in
479         to make sure we had test coverage for this branch.  I guess we do.  :)
480
481         * html/HTMLConstructionSite.cpp:
482         (WebCore::HTMLConstructionSite::attach):
483
484 2010-07-25  Pavel Feldman  <pfeldman@chromium.org>
485
486         Reviewed by Timothy Hatcher.
487
488         Web Inspector: check of jQuery object type prior to treating it as an array.
489
490         https://bugs.webkit.org/show_bug.cgi?id=42946
491
492         * inspector/front-end/InjectedScript.js:
493         (injectedScriptConstructor):
494
495 2010-07-24  Daniel Bates  <dbates@rim.com>
496
497         Unreviewed, build fix when building with WML support.
498
499         PageCache.cpp compile error when WML is enabled
500         https://bugs.webkit.org/show_bug.cgi?id=40765
501
502         Remove extraneous semicolon.
503
504         Note, building with WML support enabled is broken
505         even with this fix. See Bug #42943 for more details.
506
507         * history/PageCache.cpp:
508         (WebCore::PageCache::canCachePageContainingThisFrame):
509
510 2010-07-24  Adam Barth  <abarth@webkit.org>
511
512         Reviewed by Eric Seidel.
513
514         Atomize strings used by the HTML5 tree builder
515         https://bugs.webkit.org/show_bug.cgi?id=42875
516
517         This is about a 1% improvement to the parsing benchmark.
518
519         * html/HTMLTagNames.in:
520         * html/HTMLTreeBuilder.cpp:
521         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
522         (WebCore::HTMLTreeBuilder::processStartTag):
523         (WebCore::HTMLTreeBuilder::processEndTagForInBody):
524         (WebCore::HTMLTreeBuilder::processStartTagForInHead):
525
526 2010-07-24  Patrick Gansterer  <paroga@paroga.com>
527
528         Reviewed by Darin Adler.
529
530         Buildfix for !ENABLE(INSPECTOR) after r63891.
531         https://bugs.webkit.org/show_bug.cgi?id=42925
532
533         * bindings/js/ScriptValue.cpp: Added property svn:eol-style.
534         * bindings/js/ScriptValue.h: Added property svn:eol-style.
535         * inspector/ConsoleMessage.cpp: Added property svn:eol-style.
536         * inspector/ConsoleMessage.h: Added property svn:eol-style.
537
538 2010-07-24  Adam Barth  <abarth@webkit.org>
539
540         Reviewed by Eric Seidel.
541
542         Re-order whitespace checks in the tokenizer
543         https://bugs.webkit.org/show_bug.cgi?id=42923
544
545         This saves about 0.5% on the parsing benchmark.
546
547         * html/HTMLTokenizer.cpp:
548
549 2010-07-24  Andreas Kling  <andreas.kling@nokia.com>
550
551         Reviewed by Simon Hausmann.
552
553         [Qt] tst_QWebFrame::callQtInvokable() fails
554         https://bugs.webkit.org/show_bug.cgi?id=41065
555
556         Converting JS objects to QVariantMaps was broken.
557         This is a partial revert of <http://trac.webkit.org/changeset/61478>
558
559         * bridge/qt/qt_runtime.cpp:
560         (JSC::Bindings::convertValueToQVariant): Remove erroneous crash guard
561         around JSObject::getPropertyNames()
562
563 2010-07-24  Adam Barth  <abarth@webkit.org>
564
565         Reviewed by Eric Seidel.
566
567         Save a per-character branch in the HTML parser
568         https://bugs.webkit.org/show_bug.cgi?id=42921
569
570         The parsing benchmark claims this is a wash, but it makes me feel
571         better.
572
573         * html/HTMLToken.h:
574         (WebCore::HTMLToken::clear):
575         (WebCore::HTMLToken::beginStartTag):
576         (WebCore::HTMLToken::beginEndTag):
577         (WebCore::HTMLToken::ensureIsCharacterToken):
578         (WebCore::HTMLToken::beginComment):
579         (WebCore::HTMLToken::beginDOCTYPE):
580         * html/HTMLTokenizer.cpp:
581         (WebCore::HTMLTokenizer::bufferCharacter):
582
583 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
584
585         Reviewed by Darin Adler.
586
587         Buildfix for !ENABLE(FILTERS) after r63531.
588         https://bugs.webkit.org/show_bug.cgi?id=42914
589
590         * rendering/RenderSVGContainer.cpp: Add missing header.
591
592 2010-07-23  Kavita Kanetkar  <kkanetkar@chromium.org>
593
594         Reviewed by Joseph Pecoraro.
595
596         Implement remaining Inspector support for chrome's appcache
597         https://bugs.webkit.org/show_bug.cgi?id=42426
598
599         * inspector/InspectorApplicationCacheAgent.cpp:
600         (WebCore::InspectorApplicationCacheAgent::getApplicationCaches):
601         (WebCore::InspectorApplicationCacheAgent::buildObjectForApplicationCache):
602         (WebCore::InspectorApplicationCacheAgent::buildArrayForApplicationCacheResources):
603         (WebCore::InspectorApplicationCacheAgent::buildObjectForApplicationCacheResource):
604         * inspector/InspectorApplicationCacheAgent.h:
605         (WebCore::InspectorController::connectFrontend):
606         * loader/appcache/ApplicationCacheHost.cpp:
607         (WebCore::ApplicationCacheHost::fillResourceList):
608         (WebCore::ApplicationCacheHost::applicationCacheInfo):
609         * loader/appcache/ApplicationCacheHost.h: Changed creation + update times to double and moved appcache info here from InspectorApplicationCacheAgent.h
610         (WebCore::ApplicationCacheHost::CacheInfo::CacheInfo):
611         (WebCore::ApplicationCacheHost::ResourceInfo::ResourceInfo):
612
613 2010-07-23  Rafael Antognolli  <antognolli@profusion.mobi>
614
615         Reviewed by Antonio Gomes.
616
617         [EFL] Cleanup glib support (make it optional)
618         https://bugs.webkit.org/show_bug.cgi?id=42480
619
620         No new tests, no new features added.
621         Just add flags and directories for glib if necessary.
622         Put some ifdefs on glib calls.
623
624         * CMakeListsEfl.txt:
625         * platform/graphics/cairo/FontPlatformDataCairo.cpp:
626         (WebCore::FontPlatformData::FontPlatformData):
627
628 2010-07-23  James Robinson  <jamesr@chromium.org>
629
630         Reviewed by Dimitri Glazkov.
631
632         CanvasRenderingContext2D has lots of style errors
633         https://bugs.webkit.org/show_bug.cgi?id=42918
634
635         CanvasRenderingContext2D.h and .cpp have lots of style errors.  I'm planning to post patches
636         to both of these files soon, so to minimize merge conflicts and additional style violations
637         it seems best to fix all of these errors in bulk before posting any functional changes.  In
638         particular the header file is incorrectly indented so any correctly formatted addition will
639         be out of place.
640
641         * html/canvas/CanvasRenderingContext2D.cpp:
642         (WebCore::CanvasRenderingContext2D::CanvasRenderingContext2D):
643         (WebCore::CanvasRenderingContext2D::setFillStyle):
644         (WebCore::CanvasRenderingContext2D::transform):
645         (WebCore::CanvasRenderingContext2D::setTransform):
646         (WebCore::CanvasRenderingContext2D::arcTo):
647         (WebCore::CanvasRenderingContext2D::arc):
648         (WebCore::validateRectForCanvas):
649         (WebCore::CanvasRenderingContext2D::strokeRect):
650         (WebCore::CanvasRenderingContext2D::drawImage):
651         (WebCore::CanvasRenderingContext2D::createRadialGradient):
652         (WebCore::CanvasRenderingContext2D::createPattern):
653         (WebCore::CanvasRenderingContext2D::willDraw):
654         (WebCore::CanvasRenderingContext2D::createImageData):
655         (WebCore::CanvasRenderingContext2D::putImageData):
656         (WebCore::CanvasRenderingContext2D::setFont):
657         (WebCore::CanvasRenderingContext2D::drawTextInternal):
658         * html/canvas/CanvasRenderingContext2D.h:
659         (WebCore::CanvasRenderingContext2D::is2d):
660         (WebCore::CanvasRenderingContext2D::getLineCap):
661         (WebCore::CanvasRenderingContext2D::getLineJoin):
662         (WebCore::CanvasRenderingContext2D::state):
663         (WebCore::CanvasRenderingContext2D::):
664
665 2010-07-23  Adam Barth  <abarth@webkit.org>
666
667         Remove FIXME that we already fixed.
668
669         * html/HTMLTreeBuilder.cpp:
670         (WebCore::HTMLTreeBuilder::processCharacter):
671
672 2010-06-24  Dimitri Glazkov  <dglazkov@chromium.org>
673
674         Reviewed by Kent Tamura.
675
676         Move more form submission logic into FormSubmission.
677         https://bugs.webkit.org/show_bug.cgi?id=41110
678
679         Two major moves:
680         * Fold HTMLFormElement::prepareFormSubmission into FormSubmission::create
681         * Move non-static (non-FormData-building) methods of FormDataBuilder
682             to FormSubmission::Attributes.
683
684         Mechanical moving, no change in behavior, so no new tests.
685
686         * WebCore.xcodeproj/project.pbxproj: Gave FormDataBuilder.h "project" and
687             FormSubmission.h "private" visibility.
688         * html/HTMLFormElement.cpp:
689         (WebCore::HTMLFormElement::submit): Changed to use FormSubmission::create.
690         (WebCore::HTMLFormElement::parseMappedAttribute): Changed to use m_attributes.
691         * html/HTMLFormElement.h:
692         (WebCore::HTMLFormElement::enctype): Ditto.
693         (WebCore::HTMLFormElement::encoding): Ditto.
694         (WebCore::HTMLFormElement::acceptCharset): Ditto.
695         * loader/FormSubmission.cpp:
696         (WebCore::generateFormDataIdentifier): Moved from HTMLFormElement.
697         (WebCore::appendMailtoPostFormDataToURL): Ditto.
698         (WebCore::FormSubmission::Attributes::parseAction): Moved from FormDataBuilder.
699         (WebCore::FormSubmission::Attributes::parseEncodingType): Ditto.
700         (WebCore::FormSubmission::Attributes::parseMethodType): Ditto.
701         (WebCore::FormSubmission::FormSubmission): Made inline.
702         (WebCore::FormSubmission::create): Added logic that used to be in
703             HTMLFormElement::prepareFormSubmission.
704         * loader/FormSubmission.h:
705         (WebCore::FormSubmission::):
706         (WebCore::FormSubmission::Attributes::Attributes): Added new class.
707         (WebCore::FormSubmission::Attributes::method): Moved from FormDataBuilder.
708         (WebCore::FormSubmission::Attributes::isMultiPartForm): Ditto.
709         (WebCore::FormSubmission::Attributes::action): Ditto.
710         (WebCore::FormSubmission::Attributes::target): Ditto.
711         (WebCore::FormSubmission::Attributes::setTarget): Ditto.
712         (WebCore::FormSubmission::Attributes::encodingType): Ditto.
713         (WebCore::FormSubmission::Attributes::acceptCharset): Ditto.
714         (WebCore::FormSubmission::Attributes::setAcceptCharset): Ditto.
715         (WebCore::FormSubmission::target): Ditto.
716         (WebCore::FormSubmission::contentType): Const-ified return type.
717         (WebCore::FormSubmission::boundary): Ditto.
718         * platform/network/FormDataBuilder.cpp:
719         (WebCore::FormDataBuilder::encodingFromAcceptCharset): Renamed method for clarity,
720             made it use param instead of a class member.
721         * platform/network/FormDataBuilder.h:
722         (WebCore::FormDataBuilder::FormDataBuilder): Made constructor private.
723         * wml/WMLGoElement.cpp: Changed to use FormSubmision::Attributes.
724         * wml/WMLGoElement.h: Ditto.
725
726 2010-07-23  Adam Barth  <abarth@webkit.org>
727
728         Reviewed by Eric Seidel.
729
730         HTML5 Parser: document.write in a asynchronous script which is specified to load before page finish blows away document
731         https://bugs.webkit.org/show_bug.cgi?id=42365
732
733         This patch implements the changes Ian made to HTML5 to address these
734         issues.  Let's try it and see how well it works.
735
736         Test: http/tests/misc/write-from-dom-script.html
737
738         * dom/Document.cpp:
739         (WebCore::Document::Document):
740         (WebCore::Document::write):
741         * dom/Document.h:
742         (WebCore::Document::writeDisabled):
743         (WebCore::Document::setWriteDisabled):
744         * dom/DocumentParser.h:
745         (WebCore::DocumentParser::hasInsertionPoint):
746         * dom/ScriptElement.cpp:
747         (WebCore::ScriptElement::insertedIntoDocument):
748         (WebCore::ScriptElementData::ScriptElementData):
749         (WebCore::ScriptElementData::evaluateScript):
750         * dom/ScriptElement.h:
751         (WebCore::ScriptElementData::writeDisabled):
752         (WebCore::ScriptElementData::setWriteDisabled):
753         * html/HTMLDocumentParser.cpp:
754         (WebCore::HTMLDocumentParser::hasInsertionPoint):
755         (WebCore::HTMLDocumentParser::insert):
756         * html/HTMLDocumentParser.h:
757
758 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
759
760         Reviewed by Adam Roben.
761
762         [WINCE] Buildfix for PopupMenuWin.cpp
763         https://bugs.webkit.org/show_bug.cgi?id=42906
764
765         Use correct API for BitmapInfo landed in r63944.
766
767         * platform/win/PopupMenuWin.cpp:
768         (WebCore::PopupMenu::paint):
769
770 2010-07-23  fsamuel@chromium.org  <fsamuel@chromium.org>
771
772         Reviewed by David Hyatt.
773
774         Updated table painting to draw cell layers in the correct order.
775
776         Table Cell Layering
777         https://bugs.webkit.org/show_bug.cgi?id=40775
778
779         Tests: tables/layering/paint-test-layering-1.html
780                tables/layering/paint-test-layering-2.html
781
782         * accessibility/AccessibilityTable.cpp:
783         (WebCore::AccessibilityTable::isTableExposableThroughAccessibility):
784         (WebCore::AccessibilityTable::addChildren):
785         (WebCore::AccessibilityTable::cellForColumnAndRow):
786         * accessibility/AccessibilityTableCell.cpp:
787         (WebCore::AccessibilityTableCell::titleUIElement):
788         * accessibility/AccessibilityTableColumn.cpp:
789         (WebCore::AccessibilityTableColumn::headerObjectForSection):
790         * rendering/AutoTableLayout.cpp:
791         (WebCore::AutoTableLayout::recalcColumn):
792         * rendering/RenderTable.cpp:
793         (WebCore::RenderTable::splitColumn):
794         (WebCore::RenderTable::calcBorderLeft):
795         (WebCore::RenderTable::calcBorderRight):
796         (WebCore::RenderTable::cellAbove):
797         (WebCore::RenderTable::cellBelow):
798         (WebCore::RenderTable::cellBefore):
799         (WebCore::RenderTable::cellAfter):
800         * rendering/RenderTableSection.cpp:
801         (WebCore::RenderTableSection::RenderTableSection):
802         (WebCore::RenderTableSection::ensureRows):
803         (WebCore::RenderTableSection::addCell):
804         (WebCore::RenderTableSection::setCellWidths):
805         (WebCore::RenderTableSection::calcRowHeight):
806         (WebCore::RenderTableSection::layoutRows):
807         (WebCore::RenderTableSection::calcOuterBorderTop):
808         (WebCore::RenderTableSection::calcOuterBorderBottom):
809         (WebCore::RenderTableSection::calcOuterBorderLeft):
810         (WebCore::RenderTableSection::calcOuterBorderRight):
811         (WebCore::RenderTableSection::firstLineBoxBaseline):
812         (WebCore::compareCellPositions):
813         (WebCore::RenderTableSection::paintCell):
814         (WebCore::RenderTableSection::paintObject):
815         (WebCore::RenderTableSection::numColumns):
816         (WebCore::RenderTableSection::appendColumn):
817         (WebCore::RenderTableSection::splitColumn):
818         (WebCore::RenderTableSection::nodeAtPoint):
819         * rendering/RenderTableSection.h:
820         (WebCore::RenderTableSection::CellStruct::CellStruct):
821         (WebCore::RenderTableSection::CellStruct::primaryCell):
822         (WebCore::RenderTableSection::CellStruct::hasCells):
823         (WebCore::RenderTableSection::primaryCellAt):
824
825 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
826
827         Reviewed by Adam Roben.
828
829         [WINCE] Buildfix for Path.h
830         https://bugs.webkit.org/show_bug.cgi?id=42904
831
832         * platform/graphics/Path.h: Add missing typedef PlatformPath.
833
834 2010-07-23  Adam Barth  <abarth@webkit.org>
835
836         Reviewed by Eric Seidel.
837
838         Reserve capacity for attributes
839         https://bugs.webkit.org/show_bug.cgi?id=42877
840
841         This is about a 1.7% improvement to the parsing benchmark.
842
843         * html/HTMLToken.h:
844         (WebCore::AtomicHTMLToken::AtomicHTMLToken):
845
846 2010-07-23  Adam Barth  <abarth@webkit.org>
847
848         Reviewed by Eric Seidel.
849
850         Inline HTMLElementStack::top
851         https://bugs.webkit.org/show_bug.cgi?id=42879
852
853         According to sampling, HTMLElementStack::top is the hotest function in
854         the treebuilder.  Inlining it is about a 0.3% improvement to the
855         parsing benchmark.  This squares with how often it occurs in samples.
856
857         * html/HTMLElementStack.cpp:
858         * html/HTMLElementStack.h:
859         (WebCore::HTMLElementStack::top):
860
861 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
862
863         Reviewed by Adam Roben.
864
865         [WINCE] Buildfix for GraphicsContext
866         https://bugs.webkit.org/show_bug.cgi?id=42888
867
868         Use AffineTransform insted of TransformationMatrix and
869         add ColorSpace parameter to the drawBitmap methods.
870
871         * platform/graphics/GraphicsContext.h:
872         * platform/graphics/wince/GraphicsContextWince.cpp:
873         (WebCore::GraphicsContext::affineTransform):
874         (WebCore::GraphicsContext::fillRoundedRect):
875         (WebCore::GraphicsContext::drawBitmap):
876         (WebCore::GraphicsContext::drawBitmapPattern):
877
878 2010-07-23  Rafael Antognolli  <antognolli@profusion.mobi>
879
880         Reviewed by Antonio Gomes.
881
882         [EFL] Add support for using libcurl network backend.
883         https://bugs.webkit.org/show_bug.cgi?id=42286
884
885         Add files and flags for soup/curl specific backend being
886         used.
887         Put some ifdefs around soup/glib calls.
888
889         No new tests, no new features implemented.
890
891         * CMakeListsEfl.txt:
892         * platform/efl/FileSystemEfl.cpp:
893         (WebCore::fileSystemRepresentation):
894
895 2010-07-23  Adam Barth  <abarth@webkit.org>
896
897         Reviewed by Eric Seidel.
898
899         Add fast-reject path for InputStreamPreprocessor::peek
900         https://bugs.webkit.org/show_bug.cgi?id=42881
901
902         This is about a 2.5% speedup on the parsing benchmark.
903
904         * html/HTMLTokenizer.h:
905         (WebCore::HTMLTokenizer::InputStreamPreprocessor::peek):
906
907 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
908
909         Reviewed by Adam Roben.
910
911         [WINCE] Merge PasteboardWince.cpp into PasteboardWin.cpp
912         https://bugs.webkit.org/show_bug.cgi?id=42885
913
914         Make PasteboardWin more portable and remove the global namespace prefix
915         from GlobalLock/GlobalUnlock, since they are only macros on WinCE.
916
917         * platform/win/PasteboardWin.cpp:
918         (WebCore::PasteboardOwnerWndProc):
919         (WebCore::Pasteboard::Pasteboard):
920         (WebCore::Pasteboard::plainText):
921         (WebCore::Pasteboard::documentFragment):
922
923 2010-07-23  Justin Schuh  <jschuh@chromium.org>
924
925         Unreviewed, rolling out r63925.
926         http://trac.webkit.org/changeset/63925
927         https://bugs.webkit.org/show_bug.cgi?id=42858
928
929         Introduced a regression in history state.
930
931         * page/History.cpp:
932         (WebCore::History::urlForState):
933
934 2010-07-23  Martin Robinson  <mrobinson@igalia.com>
935
936         Unreviewed build fix.
937
938         Fix make dist by aligning the source list with currently used headers.
939
940         * GNUmakefile.am:
941
942 2010-07-23  Alex Milowski  <alex@milowski.com>
943
944         Reviewed by Darin Adler.
945
946         Moved the 'mathml' group back to where it belongs as
947         siblings of the other major folders.
948
949         * WebCore.xcodeproj/project.pbxproj:
950
951 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
952
953         Reviewed by Adam Roben.
954
955         Buildfix for WinCE graphics after r56015.
956         https://bugs.webkit.org/show_bug.cgi?id=41354
957
958         Use fabs() insetead of abs().
959
960         * platform/graphics/wince/GraphicsContextWince.cpp:
961         (WebCore::GraphicsContext::fillRect):
962         * platform/graphics/wince/PlatformPathWince.cpp:
963         (WebCore::PlatformPath::addArcTo):
964
965 2010-07-23  ryuan choi  <ryuan.choi@samsung.com>
966
967         Reviewed by Antonio Gomes.
968
969         [EFL]build break because of r63927
970         https://bugs.webkit.org/show_bug.cgi?id=42864
971
972         * CMakeLists.txt:
973
974 2010-07-23  Adam Barth  <abarth@webkit.org>
975
976         Reviewed by Eric Seidel.
977
978         Inline HTMLTokenizer::InputStreamPreprocessor::peek
979         https://bugs.webkit.org/show_bug.cgi?id=42880
980
981         According to sampling HTMLTokenizer::InputStreamPreprocessor::peek is
982         the hotest function in the tokenizer.  Inlining this function is about
983         a 2.2% speedup on the parser benchmark.  Turns out the tokenizer is
984         much hotter than the tree builder.  :)
985
986         * html/HTMLTokenizer.h:
987         (WebCore::HTMLTokenizer::InputStreamPreprocessor::peek):
988
989 2010-07-23  Andras Becsi  <abecsi@webkit.org>
990
991         Unreviewed build fix.
992
993         [Qt] Fix WebKit2 linkage.
994
995         No new tests needed.
996
997         * WebCore.pro: add InjectedBundlePageEditorClient.cpp to WK2 sources
998
999 2010-07-23  Balazs Kelemen  <kb@inf.u-szeged.hu>
1000
1001         Reviewed by Kenneth Rohde Christiansen.
1002
1003         [Qt] Setup the QtWebProcess
1004
1005         https://bugs.webkit.org/show_bug.cgi?id=42623
1006
1007         * WebCore.pro: Added WebProcessMainQt.cpp to SOURCES.
1008
1009 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1010
1011         Not reviewed: Chromium tests fix.
1012
1013         * inspector/CodeGeneratorInspector.pm:
1014
1015 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1016
1017         Reviewed by Adam Roben.
1018
1019         [WINCE] Buildfix for PluginDatabaseWin.cpp
1020         https://bugs.webkit.org/show_bug.cgi?id=41887
1021
1022         Define KEY_ENUMERATE_SUB_KEYS only if not defined.
1023
1024         * plugins/win/PluginDatabaseWin.cpp:
1025
1026 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1027
1028         Reviewed by Yury Semikhatsky.
1029
1030         Chromium DevTools: migrate from utility context-based injected
1031         script dispatch to upstream schema.
1032
1033         https://bugs.webkit.org/show_bug.cgi?id=42883
1034
1035         * bindings/v8/ScriptProfile.cpp:
1036         (WebCore::ScriptProfile::title):
1037         * inspector/InspectorController.cpp:
1038         (WebCore::InspectorController::getProfile):
1039         * inspector/front-end/InjectedScriptAccess.js:
1040
1041 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1042
1043         Not reviewed: rolling out r63963.
1044
1045         * bindings/v8/ScriptProfile.cpp:
1046         (WebCore::ScriptProfile::title):
1047         * inspector/InspectorController.cpp:
1048         (WebCore::InspectorController::getProfile):
1049         * inspector/front-end/InjectedScriptAccess.js:
1050
1051 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1052
1053         Reviewed by Yury Semikhatsky.
1054
1055         Chromium DevTools: migrate from utility context-based injected
1056         script dispatch to upstream schema.
1057
1058         https://bugs.webkit.org/show_bug.cgi?id=42883
1059
1060         * bindings/v8/ScriptProfile.cpp:
1061         (WebCore::ScriptProfile::title):
1062         * inspector/InspectorController.cpp:
1063         (WebCore::InspectorController::getProfile):
1064         * inspector/front-end/InjectedScriptAccess.js:
1065         (InjectedScriptAccess.get if):
1066         (InjectedScriptAccess):
1067
1068 2010-07-23  Yury Semikhatsky  <yurys@chromium.org>
1069
1070         Reviewed by Pavel Feldman.
1071
1072         Web Inspector: error info propagation in InspectorBackendDispatch should be improved
1073         https://bugs.webkit.org/show_bug.cgi?id=42873
1074
1075         * inspector/CodeGeneratorInspector.pm: report wrong parameter count and wrong
1076         parametere type errror to the dispatch method caller
1077
1078 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1079
1080         Reviewed by Yury Semikhatsky.
1081
1082         Web Inspector: injectedScriptId is missing in the stack frame wrappers.
1083
1084         https://bugs.webkit.org/show_bug.cgi?id=42889
1085
1086         * inspector/front-end/InjectedScript.js:
1087         (injectedScriptConstructor):
1088
1089 2010-07-23  Shinichiro Hamaji  <hamaji@chromium.org>
1090
1091         Reviewed by Kent Tamura.
1092
1093         Backslash isn't transcoded in <input>
1094         https://bugs.webkit.org/show_bug.cgi?id=42878
1095
1096         * css/CSSStyleSelector.cpp:
1097         (WebCore::CSSStyleSelector::applyProperty):
1098
1099 2010-07-22  Ilya Tikhonovsky  <loislo@chromium.org>
1100
1101         Reviewed by Yury Semikhatsky.
1102
1103         WebInspector: CodeGeneratorInspector was extended and now it can
1104         generate InspectorBackedDispatcher's code. This dispatcher will
1105         validate and dispatch incoming inspector commands only by native
1106         code without using javaScript. That is necessary step for
1107         RemoteDebugging support.
1108         https://bugs.webkit.org/show_bug.cgi?id=42588
1109
1110         * GNUmakefile.am:
1111         * WebCore.gyp/WebCore.gyp:
1112         * WebCore.gypi:
1113         * WebCore.pri:
1114         * WebCore.vcproj/WebCore.vcproj:
1115         * WebCore.xcodeproj/project.pbxproj:
1116         * inspector/CodeGeneratorInspector.pm:
1117         * inspector/InspectorController.cpp:
1118         (WebCore::InspectorController::InspectorController):
1119         * inspector/InspectorController.h:
1120         (WebCore::InspectorController::inspectorBackendDispatcher):
1121         * inspector/front-end/ElementsPanel.js:
1122         (WebInspector.ElementsPanel.prototype.setDocument):
1123
1124 2010-07-22  Shinichiro Hamaji  <hamaji@chromium.org>
1125
1126         Reviewed by Kent Tamura.
1127
1128         Backslash is transcoded into yen sign even when non japanese font is specified
1129         https://bugs.webkit.org/show_bug.cgi?id=39654
1130
1131         Test: fast/text/backslash-to-yen-sign-euc.html
1132
1133         * css/CSSStyleSelector.cpp:
1134         (WebCore::CSSStyleSelector::applyProperty): Call FontDescriptor::setIsSpecifiedFont.
1135         * platform/graphics/Font.cpp:
1136         (WebCore::Font::Font):
1137         * platform/graphics/FontDescription.h: Added m_isSpecifiedFont and its accessors.
1138         (WebCore::FontDescription::FontDescription):
1139         (WebCore::FontDescription::isSpecifiedFont):
1140         (WebCore::FontDescription::setIsSpecifiedFont):
1141         (WebCore::FontDescription::operator==):
1142         * platform/text/transcoder/FontTranscoder.cpp:
1143         (WebCore::FontTranscoder::converterType): Don't transcode backslashes if a font family is explicitly specified.
1144         (WebCore::FontTranscoder::convert):
1145         (WebCore::FontTranscoder::needsTranscoding):
1146         * platform/text/transcoder/FontTranscoder.h: Change the arguments from AtomicString fontFamily to Font.
1147         * rendering/RenderText.cpp:
1148         (WebCore::RenderText::updateNeedsTranscoding):
1149         (WebCore::RenderText::setTextInternal):
1150
1151 2010-07-22  Jon Honeycutt  <jhoneycutt@apple.com>
1152
1153         Windows build fix.
1154
1155         Commit some changes missed by the commit-queue in r63944.
1156
1157         * platform/win/BitmapInfo.h:
1158         (WebCore::BitmapInfo::is16bit):
1159         (WebCore::BitmapInfo::is32bit):
1160         (WebCore::BitmapInfo::width):
1161         (WebCore::BitmapInfo::height):
1162         (WebCore::BitmapInfo::size):
1163         (WebCore::BitmapInfo::paddedWidth):
1164         (WebCore::BitmapInfo::numPixels):
1165         (WebCore::BitmapInfo::paddedBytesPerLine):
1166         (WebCore::BitmapInfo::bytesPerLine):
1167
1168 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1169
1170         Reviewed by Adam Roben.
1171
1172         [WINCE] Add additonal methods to BitmapInfo.
1173         https://bugs.webkit.org/show_bug.cgi?id=42071
1174
1175         Added a parameter to create 16bit BitmapInfo and
1176         some additional accessors for the WinCE port.
1177
1178         Set bmiHeader.biSize only at constructor.
1179
1180         * platform/win/BitmapInfo.cpp: Added property svn:eol-style.
1181         (WebCore::bitmapInfoForSize):
1182         (WebCore::BitmapInfo::create):
1183         (WebCore::BitmapInfo::createBottomUp):
1184         * platform/win/BitmapInfo.h: Added property svn:eol-style.
1185         (WebCore::BitmapInfo::is16bit):
1186         (WebCore::BitmapInfo::is32bit):
1187         (WebCore::BitmapInfo::width):
1188         (WebCore::BitmapInfo::height):
1189         (WebCore::BitmapInfo::size):
1190         (WebCore::BitmapInfo::paddedWidth):
1191         (WebCore::BitmapInfo::numPixels):
1192         (WebCore::BitmapInfo::paddedBytesPerLine):
1193         (WebCore::BitmapInfo::bytesPerLine):
1194
1195 2010-07-22  MORITA Hajime  <morrita@google.com>
1196
1197         Reviewed by Ojan Vafai.
1198
1199         Autoscroll on selection should take scrollbars into account as outside of the content.
1200         https://bugs.webkit.org/show_bug.cgi?id=40403
1201
1202         ScrollView's scroll-bar was considered as a part of content rect,
1203         that prevents autoscroll to happen when the pointer is over the
1204         scroll-bar.  This change excluded scroll-bar region from scrolling
1205         delta calculation.
1206         
1207         Test: WebCore/manual-tests/autoscroll-over-scrollbar.html
1208
1209         * rendering/RenderLayer.cpp:
1210         (WebCore::RenderLayer::scrollRectToVisible):
1211
1212 2010-07-22  Kent Tamura  <tkent@chromium.org>
1213
1214         Reviewed by Darin Adler.
1215
1216         maxLength should not be applied to non-text types
1217         https://bugs.webkit.org/show_bug.cgi?id=42814
1218
1219         According to the HTML5 specification, maxLength attribute should
1220         be applied to text, search, url, tel, email, and password types.
1221
1222         Test: fast/forms/input-maxlength-unsupported.html
1223
1224         * dom/InputElement.cpp:
1225         (WebCore::InputElement::handleBeforeTextInsertedEvent):
1226          Apply maxLength() only if supportsMaxLength() is true.
1227         * dom/InputElement.h:
1228         * html/HTMLInputElement.cpp:
1229         (WebCore::HTMLInputElement::patternMismatch):
1230          Code cleanup by isTextType().
1231         (WebCore::HTMLInputElement::tooLong):
1232          ditto.
1233         (WebCore::HTMLInputElement::isTextType):
1234          Added. This returns true for types with maxLength/pattern/placeholder.
1235         * html/HTMLInputElement.h:
1236         (WebCore::HTMLInputElement::supportsMaxLength):
1237          Added. Just calls isTextType().
1238         (WebCore::HTMLInputElement::supportsPlaceholder):
1239          Code cleanup by isTextType().
1240         * wml/WMLInputElement.h:
1241         (WebCore::WMLInputElement::supportsMaxLength):
1242          Added. Always returns true.
1243
1244 2010-07-22  Ryosuke Niwa  <rniwa@webkit.org>
1245
1246         Reviewed by Kent Tamura.
1247
1248         [cleanup] logic to extract adjacent lists and list children in listifyParagraph and unlistifyParagraph should be isolated
1249         https://bugs.webkit.org/show_bug.cgi?id=42841
1250
1251         Added enclosingListChild that takes the list node containing the node for the second argument
1252         to ensure that it only returns the enclosing list child of the specified list node.
1253
1254         Added adjacentEnclosingList that finds adjacent lists of the specified position.
1255         It takes the current position, the adjacent position that may or may not have a list, and the list type,
1256         and avoids a list that belongs to a different table cell or an outer list that contains the current position.
1257
1258         No new tests needed.
1259
1260         * editing/InsertListCommand.cpp:
1261         (WebCore::enclosingListChild): Added. See above.
1262         (WebCore::InsertListCommand::unlistifyParagraph): Uses enclosingListChild.
1263         (WebCore::adjacentEnclosingList): Added. See above.
1264         (WebCore::InsertListCommand::listifyParagraph): Uses adjacentEnclosingList.
1265
1266 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1267
1268         Reviewed by Adam Roben.
1269
1270         [WIN] Correct DeleteObject usage in PasteboardWin
1271         https://bugs.webkit.org/show_bug.cgi?id=41956
1272
1273         Call DeleteObject() if OpenClipboard() fails and don't
1274         call DeleteObject() on objects returned from SelectObject().
1275
1276         * platform/win/PasteboardWin.cpp:
1277         (WebCore::Pasteboard::writeImage):
1278
1279 2010-07-22  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
1280
1281         Reviewed by Antonio Gomes.
1282
1283         [EFL] Log function is incorrectly used.
1284         r63804 used LOG function, but it should receive one more parameter.
1285         Used LOG_ERROR instead, since this is an error message.
1286         https://bugs.webkit.org/show_bug.cgi?id=42842
1287
1288         * platform/efl/WidgetEfl.cpp:
1289         (WebCore::Widget::applyFallbackCursor): use LOG_ERROR in place of LOG
1290
1291 2010-07-22  Kent Tamura  <tkent@chromium.org>
1292
1293         Reviewed by Ojan Vafai.
1294
1295         Cancel button of search input field should fire input/change events
1296         https://bugs.webkit.org/show_bug.cgi?id=34393
1297
1298         Test: fast/forms/search-cancel-button-events.html
1299
1300         * rendering/TextControlInnerElements.cpp:
1301         (WebCore::SearchFieldCancelButtonElement::defaultEventHandler):
1302
1303 2010-07-22  Sam Weinig  <sam@webkit.org>
1304
1305         Reviewed by Maciej Stachowiak.
1306
1307         Fix for <rdar://problem/8222626>
1308         Send textDidChangeInTextField delegate callback only in response to typing or other forms of user text input.
1309
1310         * GNUmakefile.am:
1311         * WebCore.gypi:
1312         * WebCore.pro:
1313         * WebCore.vcproj/WebCore.vcproj:
1314         * WebCore.xcodeproj/project.pbxproj:
1315         Update project files.
1316         
1317         * dom/UserTypingGestureIndicator.cpp: Copied from dom/UserGestureIndicator.cpp.
1318         (WebCore::UserTypingGestureIndicator::processingUserTypingGesture):
1319         (WebCore::focusedNode):
1320         (WebCore::UserTypingGestureIndicator::focusedElementAtGestureStart):
1321         (WebCore::UserTypingGestureIndicator::UserTypingGestureIndicator):
1322         (WebCore::UserTypingGestureIndicator::~UserTypingGestureIndicator):
1323         * dom/UserTypingGestureIndicator.h: Copied from dom/UserGestureIndicator.h.
1324         Version of UserGestureIndicator that just tracks keyboard gestures and the focused
1325         node they were targeting.
1326         
1327         * editing/Editor.cpp:
1328         (WebCore::Editor::confirmComposition):
1329         (WebCore::Editor::setComposition):
1330         * page/EventHandler.cpp:
1331         (WebCore::EventHandler::keyEvent):
1332         Put UserTypingGestureIndicator on the stack for user generated keyboard events.
1333
1334 2010-07-22  Kent Tamura  <tkent@chromium.org>
1335
1336         Reviewed by Ojan Vafai.
1337
1338         <input type=number> stepper buttons should dispatch input/change events
1339         https://bugs.webkit.org/show_bug.cgi?id=42440
1340
1341         Test: fast/forms/input-number-events.html
1342
1343         * html/HTMLInputElement.cpp:
1344         (WebCore::HTMLInputElement::stepUpFromRenderer):
1345          Sets a flag to dispatch 'change' event and dispatches 'input' event
1346          if the value is changed.
1347
1348 2010-07-22  Justin Schuh  <jschuh@chromium.org>
1349
1350         Reviewed by Adam Barth.
1351
1352         Make history state methods validate against document URL
1353         https://bugs.webkit.org/show_bug.cgi?id=42858
1354
1355         Test: fast/history/history-replace-illegal.html
1356
1357         * page/History.cpp:
1358         (WebCore::History::urlForState):
1359
1360 2010-07-22  Andy Estes  <aestes@apple.com>
1361
1362         Reviewed by Maciej Stachowiak.
1363
1364         When there are no beforeload listeners on a document,
1365         HTMLLinkElement::process() should be called immediately when the node is
1366         inserted into the document, rather than waiting until after attach is
1367         performed.
1368         <https://bugs.webkit.org/show_bug.cgi?id=42859>
1369         <rdar://problem/8194528>
1370
1371         No new tests.
1372
1373         * html/HTMLLinkElement.cpp:
1374         (WebCore::HTMLLinkElement::insertedIntoDocument): If there are no
1375         beforeload listeners, call process() immediately.  Otherwise, add it as
1376         a post-attach callback.
1377
1378 2010-07-22  James Robinson  <jamesr@chromium.org>
1379
1380         Reviewed by Simon Fraser.
1381
1382         REGRESSION: box shadows on WebGL canvas elements repainted every frame
1383         https://bugs.webkit.org/show_bug.cgi?id=42852
1384
1385         Avoid calling HTMLCanvasElement::willDraw() every time a WebGL canvas'
1386         contents update.  This call is only needed in the non-compositing
1387         rendering path.
1388
1389         Covered by fast/canvas/webgl/canvas-test.html
1390
1391         * html/canvas/WebGLRenderingContext.cpp:
1392         (WebCore::WebGLRenderingContext::markContextChanged):
1393         (WebCore::WebGLRenderingContext::paintRenderingResultsToCanvas):
1394
1395 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1396
1397         Reviewed by Ojan Vafai.
1398
1399         Remove useless PlatformScrollBar
1400         https://bugs.webkit.org/show_bug.cgi?id=41911
1401
1402         * WebCore.gypi:
1403         * platform/win/PlatformScrollBar.h: Removed.
1404         * platform/win/PlatformScrollBarWin.cpp: Removed.
1405
1406 2010-07-22  Ryosuke Niwa  <rniwa@webkit.org>
1407
1408         Reviewed by Ojan Vafai.
1409
1410         [Windows] Home hey doesn't work in first DIV inside a TABLE
1411         https://bugs.webkit.org/show_bug.cgi?id=42767
1412
1413         The bug was caused by positionAvoidingFirstPositionInTable moving outside of the editing region.
1414         Added a check so that it avoids the first position in the table only if the table is editable.
1415
1416         Test: editing/selection/home-inside-noneditable-table.html
1417
1418         * editing/visible_units.cpp:
1419         (WebCore::positionAvoidingFirstPositionInTable):
1420
1421 2010-07-22  Antonio Gomes  <tonikitoo@webkit.org>
1422
1423         Reviewed by Laszlo Gombos.
1424
1425         Fix build warning for missing parentheses.
1426
1427         WebCore/html/HTMLInputElement.cpp:113: warning: suggest parentheses around '&&' within '||'
1428
1429         * html/HTMLInputElement.cpp:
1430         (WebCore::isNumberCharacter):
1431
1432 2010-07-22  John Gregg  <johnnyg@google.com>
1433
1434         Reviewed by Kent Tamura.
1435
1436         [chromium] fast/forms/input-file-directory-upload.html fails on win after r63454
1437         https://bugs.webkit.org/show_bug.cgi?id=42768
1438
1439         Normalize path separators for the script API by convering \'s to /'s.
1440
1441         Covered by input-file-directory-upload.html
1442
1443         * html/HTMLInputElement.cpp:
1444         (WebCore::HTMLInputElement::setFileListFromRenderer):
1445
1446 2010-07-22  Dmitry Titov  <dimich@chromium.org>
1447
1448         Rubberstamped by Simon Fraser.
1449
1450         * platform/PlatformStrategies.h:
1451         (WebCore::PlatformStrategies::PlatformStrategies): Add initialization of m_localizationStrategy.
1452
1453 2010-07-22  Dumitru Daniliuc  <dumi@chromium.org>
1454
1455         Reviewed by Ojan Vafai.
1456
1457         Make db.changeVersion()'s third parameter optional, as required by the spec.
1458         https://bugs.webkit.org/show_bug.cgi?id=42787
1459
1460         * bindings/js/JSDatabaseCustom.cpp:
1461         (WebCore::JSDatabase::changeVersion):
1462         * bindings/js/JSDatabaseSyncCustom.cpp:
1463         (WebCore::JSDatabaseSync::changeVersion):
1464         * bindings/v8/custom/V8DatabaseCustom.cpp:
1465         (WebCore::V8Database::changeVersionCallback):
1466         * bindings/v8/custom/V8DatabaseSyncCustom.cpp:
1467         (WebCore::V8DatabaseSync::changeVersionCallback):
1468         * storage/SQLTransaction.cpp:
1469         (WebCore::SQLTransaction::deliverTransactionCallback):
1470         * storage/SQLTransactionSync.cpp:
1471         (WebCore::SQLTransactionSync::execute):
1472
1473 2010-07-20  Jessie Berlin  <jberlin@apple.com>
1474
1475         Reviewed by Darin Adler.
1476
1477         Code Generator: Allow negative and string constants.
1478         Also add tests for hexadecimal numbers.
1479
1480         * bindings/scripts/CodeGeneratorJS.pm:
1481         If it is a string, do not try to cast it to an int.
1482
1483         * bindings/scripts/IDLStructure.pm:
1484         Add a regex just for constant values that allows strings, hexadecimal numbers, and integral numbers.
1485
1486         * bindings/scripts/test/CPP/WebDOMTestObj.h:
1487         Update the tests.
1488
1489         * bindings/scripts/test/JS/JSTestObj.cpp:
1490         Ditto.
1491
1492         * bindings/scripts/test/JS/JSTestObj.h:
1493         Ditto.
1494
1495         * bindings/scripts/test/ObjC/DOMTestObj.h:
1496         Ditto.
1497
1498         * bindings/scripts/test/TestObj.idl:
1499         Add tests for negative numbers, strings, and hexadecimal numbers.
1500
1501         * bindings/scripts/test/V8/V8TestObj.cpp:
1502         Update the tests.
1503
1504 2010-07-22  Simon Fraser  <simon.fraser@apple.com>
1505
1506         Reviewed by Darin Adler.
1507
1508         Composited layers don't scroll in WebKit2
1509         https://bugs.webkit.org/show_bug.cgi?id=42771
1510
1511         More prep work: make a single scrollPositionChanged() bottleneck method, and use it to 
1512         send the scroll event, repaint fixed elements, and update compositing layers.
1513
1514         * page/FrameView.cpp:
1515         (WebCore::FrameView::scrollPositionChangedViaPlatformWidget):
1516         (WebCore::FrameView::scrollPositionChanged):
1517         (WebCore::FrameView::repaintFixedElementsAfterScrolling):
1518         (WebCore::FrameView::valueChanged):
1519         * page/FrameView.h:
1520         * platform/ScrollView.cpp:
1521         (WebCore::ScrollView::valueChanged):
1522
1523 2010-07-22  Sam Weinig  <sam@webkit.org>
1524
1525         Reviewed by Anders Carlsson.
1526
1527         Patch for https://bugs.webkit.org/show_bug.cgi?id=42836
1528         Add localized strings stubs for WebKit2.
1529
1530         - Convert WebKit (mac) to use a LocalizationStrategy rather than the WebCoreViewFactory
1531           for localized strings.
1532
1533         Gets us below 200 tests failing in WebKitTestRunner.
1534
1535         * WebCore.exp.in:
1536         * WebCore.xcodeproj/project.pbxproj:
1537         * page/mac/WebCoreViewFactory.h:
1538         * platform/LocalizationStrategy.h: Added.
1539         (WebCore::LocalizationStrategy::~LocalizationStrategy):
1540         * platform/PlatformStrategies.h:
1541         (WebCore::PlatformStrategies::pluginStrategy):
1542         (WebCore::PlatformStrategies::localizationStrategy):
1543         (WebCore::PlatformStrategies::PlatformStrategies):
1544         (WebCore::PlatformStrategies::~PlatformStrategies):
1545         * platform/mac/LocalizedStringsMac.mm:
1546
1547 2010-07-22  Andras Becsi  <abecsi@webkit.org>
1548
1549         Reviewed by Antonio Gomes.
1550
1551         [Qt] Implement WebContext::applicationCacheDirectory() for the Qt port of WebKit2.
1552         https://bugs.webkit.org/show_bug.cgi?id=42830
1553
1554         No new tests needed.
1555
1556         * WebCore.pri: add ApplicationCacheStorage.h to wk2 headers
1557         * WebCore.pro: add WebContextQt.cpp to sources
1558
1559 2010-07-22  Joone Hur  <joone.hur@samsung.com>
1560
1561         Reviewed by Antonio Gomes.
1562
1563         [EFL] add support for logging
1564         https://bugs.webkit.org/show_bug.cgi?id=40807
1565
1566         Enable to turn log channels on for EFL port
1567
1568         * platform/efl/LoggingEfl.cpp:
1569         (WebCore::InitializeLoggingChannelsIfNecessary):
1570
1571 2010-07-22  Balazs Kelemen  <kb@inf.u-szeged.hu>
1572
1573         Ureviewed build fix.
1574
1575         [Qt] Fix WebKit2 build
1576
1577         https://bugs.webkit.org/show_bug.cgi?id=42638
1578
1579         * WebCore.pri:
1580         * WebCore.pro:
1581
1582 2010-07-22  Andrey Kosyakov  <caseq@chromium.org>
1583
1584         Reviewed by Yury Semikhatsky.
1585
1586         Web Inspector: Stop granting universal access to inspector front-end page.
1587         https://bugs.webkit.org/show_bug.cgi?id=42817
1588
1589         * inspector/InspectorFrontendClientLocal.cpp:
1590         (WebCore::InspectorFrontendClientLocal::windowObjectCleared):
1591
1592 2010-07-22  Yury Semikhatsky  <yurys@chromium.org>
1593
1594         Unreviewed. Fix inspector console tests crashes.
1595
1596         * bindings/v8/ScriptValue.cpp:
1597         (WebCore::v8ToInspectorValue):
1598         (WebCore::ScriptValue::toInspectorValue): enter v8 context before calling GetPropertyNames
1599
1600 2010-07-22  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1601
1602         Unreviewed build fix.
1603
1604         [EFL] Build break for SVN r63881
1605         https://bugs.webkit.org/show_bug.cgi?id=42813
1606
1607         * CMakeLists.txt: Add platform/SchemeRegistry.cpp
1608
1609 2010-07-22  Hayato Ito  <hayato@chromium.org>
1610
1611         Reviewed by Darin Adler.
1612
1613         Refactor CSSSelector's destructor and make it inline.
1614
1615         https://bugs.webkit.org/show_bug.cgi?id=42726
1616
1617         Refactor with no behavior change, thus no new tests.
1618
1619         * css/CSSSelector.cpp:
1620         (WebCore::CSSSelectorBag::~CSSSelectorBag): Make sure the bag is empty.
1621         (WebCore::CSSSelectorBag::add): Renamed from append().
1622         (WebCore::CSSSelector::releaseOwnedSelectorsToBag):
1623         (WebCore::CSSSelector::deleteReachableSelectors):
1624         * css/CSSSelector.h:
1625         (WebCore::CSSSelector::~CSSSelector):
1626
1627 2010-07-22  Yury Semikhatsky  <yurys@chromium.org>
1628
1629         Reviewed by Pavel Feldman.
1630
1631         Web Inspector: should be possible to convert console message arguments to InspectorValues
1632         https://bugs.webkit.org/show_bug.cgi?id=42457
1633
1634         Now it is possible to convert simple JavaScript objects to
1635         InspectorValues.
1636
1637         * bindings/js/ScriptValue.cpp:
1638         (WebCore::jsToInspectorValue):
1639         (WebCore::ScriptValue::toInspectorValue):
1640         * bindings/js/ScriptValue.h:
1641         * bindings/v8/ScriptValue.cpp:
1642         (WebCore::v8ToInspectorValue):
1643         (WebCore::ScriptValue::toInspectorValue):
1644         * bindings/v8/ScriptValue.h:
1645         * inspector/ConsoleMessage.cpp: console notifications are pushed to RemoteInspectorFrontend instead of InspectorFrontend.
1646         (WebCore::ConsoleMessage::CallFrame::buildInspectorObject):
1647         (WebCore::ConsoleMessage::addToFrontend):
1648         (WebCore::ConsoleMessage::updateRepeatCountInConsole):
1649         * inspector/ConsoleMessage.h:
1650         * inspector/InjectedScript.cpp:
1651         (WebCore::InjectedScript::wrapForConsole): return InspectorValue instead of SerializedScriptValue
1652         * inspector/InjectedScript.h:
1653         * inspector/Inspector.idl:
1654         * inspector/InspectorBackend.cpp:
1655         (WebCore::InspectorBackend::clearConsoleMessages): send response directly from the backend
1656         (WebCore::InspectorBackend::remoteFrontend):
1657         * inspector/InspectorBackend.h:
1658         * inspector/InspectorController.cpp:
1659         (WebCore::InspectorController::addConsoleMessage):
1660         (WebCore::InspectorController::clearConsoleMessages): don't send notification to the front end,
1661         it will be send either from InspectorBackend.cpp if it was a user request or reset message will
1662         be send to the front end if the messages are cleared due to page navigation.
1663         (WebCore::InspectorController::populateScriptObjects):
1664         * inspector/front-end/ConsoleView.js:
1665         * inspector/front-end/inspector.js:
1666
1667 2010-07-22  Alexander Pavlov  <apavlov@chromium.org>
1668
1669         Reviewed by Pavel Feldman.
1670
1671         Web Inspector: Modifying settings actually modifies defaultValues
1672         https://bugs.webkit.org/show_bug.cgi?id=42816
1673
1674         Avoid using live modifiable objects as property default values.
1675
1676         * inspector/front-end/Settings.js:
1677         (WebInspector.Settings):
1678         (WebInspector.Settings.prototype.reset):
1679         (WebInspector.Settings.prototype._load):
1680         (WebInspector.Settings.prototype.installSetting):
1681         (WebInspector.Settings.prototype._get):
1682
1683 2010-07-22  Ben Murdoch  <benm@google.com>
1684
1685         Reviewed by Simon Fraser.
1686
1687         Touch events do not affect the :active CSS state
1688         https://bugs.webkit.org/show_bug.cgi?id=39493
1689
1690         Clean the code up by using a typedef based on comments
1691         post landing of the original patch.
1692
1693         No change in functionality so no new tests.
1694
1695         * rendering/HitTestRequest.h: Typedef the HitTestRequestType to unsigned.
1696         * WebCore.exp.in: Update export signature.
1697         * WebCore.order: Ditto.
1698         * page/EventHandler.cpp: use the new typedef.
1699         * page/EventHandler.h: ditto.
1700
1701
1702 2010-07-22  Alexander Pavlov  <apavlov@chromium.org>
1703
1704         Reviewed by Pavel Feldman.
1705
1706         Web Inspector: Rename method for CSS rule source range retrieval and fix return object format
1707         https://bugs.webkit.org/show_bug.cgi?id=42379
1708
1709         * inspector/Inspector.idl:
1710         * inspector/InspectorBackend.cpp:
1711         (WebCore::InspectorBackend::getRuleRanges):
1712         * inspector/InspectorBackend.h:
1713         * inspector/InspectorBackend.idl:
1714         * inspector/InspectorCSSStore.cpp:
1715         (WebCore::InspectorCSSStore::getRuleRanges):
1716         * inspector/InspectorCSSStore.h:
1717         * inspector/InspectorDOMAgent.cpp:
1718         (WebCore::InspectorDOMAgent::getRuleRanges):
1719         * inspector/InspectorDOMAgent.h:
1720         * inspector/front-end/DOMAgent.js:
1721
1722 2010-07-21  Pavel Feldman  <pfeldman@chromium.org>
1723
1724         Reviewed by Yury Semikhatsky.
1725
1726         Chromium Dev Tools: Separators are missing in the context menu.
1727
1728         https://bugs.webkit.org/show_bug.cgi?id=42761
1729
1730         * bindings/js/JSInspectorFrontendHostCustom.cpp:
1731         (WebCore::JSInspectorFrontendHost::showContextMenu):
1732         * bindings/v8/custom/V8InspectorFrontendHostCustom.cpp:
1733         (WebCore::V8InspectorFrontendHost::showContextMenuCallback):
1734         * platform/ContextMenuItem.h:
1735         (WebCore::):
1736
1737 2010-07-22  Zoltan Herczeg  <zherczeg@webkit.org>
1738
1739         Reviewed by Nikolas Zimmermann.
1740
1741         SVGFilterElement & SVGFE*Element don't support dynamic invalidation, when attributes change
1742         https://bugs.webkit.org/show_bug.cgi?id=42244
1743
1744         Implementing svgAttributeChanged and childrenChanged for the base class of all filters,
1745         called SVGFilterPrimitiveStandardAttributes.
1746
1747         Tests: svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-height-attr.html
1748                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-result-attr.html
1749                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-width-attr.html
1750                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-x-attr.html
1751                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-y-attr.html
1752                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-height-prop.html
1753                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-result-prop.html
1754                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-width-prop.html
1755                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-x-prop.html
1756                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-y-prop.html
1757
1758         * svg/SVGFilterPrimitiveStandardAttributes.cpp:
1759         (WebCore::SVGFilterPrimitiveStandardAttributes::svgAttributeChanged):
1760         (WebCore::SVGFilterPrimitiveStandardAttributes::childrenChanged):
1761         * svg/SVGFilterPrimitiveStandardAttributes.h:
1762         (WebCore::SVGFilterPrimitiveStandardAttributes::invalidateFilter):
1763
1764 2010-07-22  Kent Tamura  <tkent@chromium.org>
1765
1766         Reviewed by Darin Adler.
1767
1768         Small refactoring for input value sanitization
1769         https://bugs.webkit.org/show_bug.cgi?id=42807
1770
1771         Rename some functions to clarify their roles.
1772
1773         * dom/InputElement.cpp:
1774         (WebCore::replaceEOLAndLimitLength):
1775           Renamed from sanitizeUserInputValue().
1776         (WebCore::InputElement::sanitizeValueForTextField):
1777           Renamed from sanitizeValue(), and call replaceEOLAndLimitLength()
1778           instead of sanitizeUserInputValue().
1779         (WebCore::InputElement::sanitizeUserInputValue):
1780           Just call replaceEOLAndLimitLength().
1781         (WebCore::InputElement::updateValueIfNeeded):
1782           Use non-static sanitizeValue() for consistency.
1783         * dom/InputElement.h:
1784         * html/HTMLInputElement.cpp:
1785         (WebCore::HTMLInputElement::setInputType):
1786           Revert the change of r63876. It is not needed because of the
1787           updateValueIfNeeded() change.
1788         (WebCore::HTMLInputElement::sanitizeValue):
1789           Apply the sanitizeValue() renaming.
1790
1791 2010-07-22  Tony Gentilcore  <tonyg@chromium.org>
1792
1793         Reviewed by Darin Fisher.
1794
1795         webkitPerformance.timing.responseEnd should not include document parse time
1796         https://bugs.webkit.org/show_bug.cgi?id=42797
1797
1798         No new tests because timing based test would be flaky.
1799
1800         * loader/FrameLoader.cpp:
1801         (WebCore::FrameLoader::finishedLoading): finishedLoading() is called by the platform at the right time, but didReceiveData() synchronously invokes parsing without returning to the event loop prior to this. So by the time the didFinishLoading() method is executed, parsing is finished. The solution is to move this time to didReceiveData() prior to parsing.
1802         * loader/MainResourceLoader.cpp:
1803         (WebCore::MainResourceLoader::didReceiveData): Update the time each time didReceiveData() is called.
1804         (WebCore::MainResourceLoader::didFinishLoading): When finished, and after parsing, set responseEnd appropriately.
1805         * loader/MainResourceLoader.h:
1806         * page/Timing.cpp:
1807         (WebCore::Timing::resourceLoadTimeRelativeToAbsolute): Now that responseEnd is set correctly, it can be much shorter for very fast loads (like local files). This exposed the fact that this skew check was not tight enough. We need to make sure that the whole range of values from the ResourceLoadTiming API fit within fetchStart-responseEnd, not just that requestTime fits in this range.
1808
1809 2010-07-21  Kent Tamura  <tkent@chromium.org>
1810
1811         Reviewed by Darin Adler.
1812
1813         Assertion failure by changing the type of an input element with a
1814         non-number value to 'range'.
1815         https://bugs.webkit.org/show_bug.cgi?id=42643
1816
1817         Test: fast/forms/input-value-sanitization.html
1818
1819         * html/HTMLInputElement.cpp:
1820         (WebCore::HTMLInputElement::setInputType):
1821         Update the value by HTMLInputElement::sanitizeValue() in a case that
1822         storesValueSeparateFromAttribute() state is not changed.
1823
1824 2010-07-21  Adam Barth  <abarth@webkit.org>
1825
1826         Reviewed by Eric Seidel.
1827
1828         Implement spec changes for basefont and bgsound
1829         https://bugs.webkit.org/show_bug.cgi?id=42792
1830
1831         * html/HTMLTreeBuilder.cpp:
1832         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
1833         (WebCore::HTMLTreeBuilder::processStartTag):
1834         (WebCore::HTMLTreeBuilder::processStartTagForInHead):
1835
1836 2010-07-21  Adam Barth  <abarth@webkit.org>
1837
1838         Reviewed by Eric Seidel.
1839
1840         Update a list of tags to match changes in the HTML5 spec
1841         https://bugs.webkit.org/show_bug.cgi?id=42791
1842
1843         We need to atomize these strings at some point.
1844
1845         * html/HTMLTreeBuilder.cpp:
1846         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
1847
1848 2010-07-21  Adam Barth  <abarth@webkit.org>
1849
1850         Reviewed by Eric Seidel.
1851
1852         HTMLTreeBuilder needs to update to match new spec behavior for <button>
1853         https://bugs.webkit.org/show_bug.cgi?id=42233
1854
1855         The spec has some typos in this area, but this is my best guess for
1856         what Ian means.  I've filed bugs against the spec for each typo.
1857
1858         * html/HTMLElementStack.cpp:
1859         (WebCore::HTMLNames::isScopeMarker):
1860         (WebCore::HTMLNames::isButtonScopeMarker):
1861         (WebCore::HTMLElementStack::inButtonScope):
1862         * html/HTMLElementStack.h:
1863         * html/HTMLTreeBuilder.cpp:
1864         (WebCore::HTMLTreeBuilder::processFakePEndTagIfPInButtonScope):
1865         (WebCore::HTMLTreeBuilder::processCloseWhenNestedTag):
1866         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
1867         (WebCore::HTMLTreeBuilder::processEndTagForInBody):
1868         * html/HTMLTreeBuilder.h:
1869
1870 2010-07-21  Justin Schuh  <jschuh@chromium.org>
1871
1872         Reviewed by Oliver Hunt.
1873
1874         Prevent DeleteButtonController enable state from changing when not editing
1875         https://bugs.webkit.org/show_bug.cgi?id=42659
1876
1877         Test: svg/custom/use-invalid-html.xhtml
1878
1879         * dom/ContainerNode.cpp:
1880         (WebCore::ContainerNode::cloneChildNodes):
1881
1882 2010-07-21  Beth Dakin  <bdakin@apple.com>
1883
1884         Reviewed by Dan Bernstein.
1885
1886         Fix for https://bugs.webkit.org/show_bug.cgi?id=42605 New border-
1887         radius path-based drawing code has some issues with corner-joins
1888
1889         GraphicsContext::clipConvexPolygon() now takes a boolean parameter 
1890         indicating whether or not the clip should be antialiased.
1891         * platform/graphics/GraphicsContext.h:
1892         * platform/graphics/cairo/GraphicsContextCairo.cpp:
1893         (WebCore::GraphicsContext::clipConvexPolygon):
1894         * platform/graphics/cg/GraphicsContextCG.cpp:
1895         (WebCore::GraphicsContext::clipConvexPolygon):
1896         * platform/graphics/haiku/GraphicsContextHaiku.cpp:
1897         (WebCore::GraphicsContext::clipConvexPolygon):
1898         * platform/graphics/qt/GraphicsContextQt.cpp:
1899         (WebCore::GraphicsContext::clipConvexPolygon):
1900         * platform/graphics/skia/GraphicsContextSkia.cpp:
1901         (WebCore::GraphicsContext::clipConvexPolygon):
1902         * platform/graphics/wince/GraphicsContextWince.cpp:
1903         (WebCore::GraphicsContext::clipConvexPolygon):
1904         * platform/graphics/wx/GraphicsContextWx.cpp:
1905         (WebCore::GraphicsContext::clipConvexPolygon):
1906
1907         Compute 4 booleans determining if each edge is a "matching" edge of 
1908         not. Send the two appropriate values to clipBorderSidePolygon() so 
1909         we can decide there whether or not to antialias the clip based on 
1910         matchy-ness.
1911         * rendering/RenderBoxModelObject.cpp:
1912         (WebCore::RenderBoxModelObject::paintBorder):
1913
1914         If both edges are matching, don't antialias. If neither edge 
1915         matches, never antialias. If one edge matches and one does not, 
1916         apply two clips to get the appropriate antialiasing in each corner.
1917         (WebCore::RenderBoxModelObject::clipBorderSidePolygon):
1918         * rendering/RenderBoxModelObject.h:
1919
1920 2010-07-21  Brady Eidson  <beidson@apple.com>
1921
1922         Reviewed by Geoffrey Garen.
1923
1924         Break out "scheme registration" functionality from SecurityOrigin to a SchemeRegistry
1925         https://bugs.webkit.org/show_bug.cgi?id=42783
1926
1927         No new tests. (No change in behavior)
1928
1929         * GNUmakefile.am:
1930         * WebCore.exp.in:
1931         * WebCore.gypi:
1932         * WebCore.pro:
1933         * WebCore.vcproj/WebCore.vcproj:
1934         * WebCore.xcodeproj/project.pbxproj:
1935         
1936         * loader/FrameLoader.cpp:
1937         (WebCore::FrameLoader::isMixedContent):
1938         (WebCore::FrameLoader::loadFrameRequest):
1939         
1940         * page/SecurityOrigin.cpp:
1941         (WebCore::SecurityOrigin::SecurityOrigin):
1942         (WebCore::SecurityOrigin::canLoad):
1943         (WebCore::SecurityOrigin::isLocal):
1944         * page/SecurityOrigin.h:
1945         
1946         * platform/SchemeRegistry.cpp: Added.
1947         (WebCore::localSchemes):
1948         (WebCore::secureSchemes):
1949         (WebCore::schemesWithUniqueOrigins):
1950         (WebCore::SchemeRegistry::registerURLSchemeAsLocal):
1951         (WebCore::SchemeRegistry::removeURLSchemeRegisteredAsLocal):
1952         (WebCore::SchemeRegistry::localURLSchemes):
1953         (WebCore::SchemeRegistry::shouldTreatURLAsLocal):
1954         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsLocal):
1955         (WebCore::SchemeRegistry::registerURLSchemeAsNoAccess):
1956         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsNoAccess):
1957         (WebCore::SchemeRegistry::registerURLSchemeAsSecure):
1958         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsSecure):
1959         * platform/SchemeRegistry.h: Added.
1960
1961 2010-07-21  Chris Marrin  <cmarrin@apple.com>
1962
1963         Reviewed by Simon Fraser.
1964
1965         Assertion failure in AnimationBase::updateStateMachine() coming out of paused state
1966         https://bugs.webkit.org/show_bug.cgi?id=37993
1967         
1968         Added logic to properly handle pausing and resuming when in the 
1969         AnimationStateStartWaitStyleAvailable state. This was causing an
1970         assert when going out of the pause state because the paused flag
1971         was not set.
1972         
1973         The fix is a straightforward implementation, going into a new
1974         AnimationStatePausedWaitStyleAvailable state and setting the paused
1975         flag (actually setting the m_pauseTime variable to something other
1976         than -1). Also added handling of the new state, both when the 
1977         "style available" callback comes in while in this state and when
1978         unpausing while in this state. 
1979         
1980         For now a LayoutTest is not possible since there's no way to go in
1981         and out of the pause state. I've opened https://bugs.webkit.org/show_bug.cgi?id=42790
1982         to track this.
1983
1984         * page/animation/AnimationBase.cpp:
1985         (WebCore::AnimationBase::updateStateMachine):
1986         * page/animation/AnimationBase.h:
1987         (WebCore::AnimationBase::):
1988
1989 2010-07-21  Bo Liu  <boliu@chromium.org>
1990
1991         Reviewed by Darin Fisher.
1992
1993         [Chromium] Add chromium WebMediaPlayer to PlatformMedia
1994         https://bugs.webkit.org/show_bug.cgi?id=41295
1995
1996         * platform/graphics/MediaPlayer.h:
1997         (WebCore::PlatformMedia::):
1998
1999 2010-07-21  Simon Fraser  <simon.fraser@apple.com>
2000
2001         Reviewed by Anders Carlsson.
2002
2003         Composited layers don't scroll in WebKit2
2004         https://bugs.webkit.org/show_bug.cgi?id=42771
2005
2006         Prep work: FrameView::scrollPositionChanged() sounds like a generic "did scroll" bottleneck,
2007         but this is deceiving. It's only every called on one platform (Mac) when the NSScrollView gets
2008         scrolled, so rename it to FrameView::scrollPositionChangedViaPlatformWidget().
2009
2010         * WebCore.exp.in:
2011         * page/FrameView.cpp:
2012         (WebCore::FrameView::scrollPositionChangedViaPlatformWidget):
2013         * page/FrameView.h:
2014
2015 2010-07-21  Peter Beverloo  <peter@lvp-media.com>
2016
2017         Reviewed by Eric Carlson.
2018
2019         Removed support for the -khtml CSS vendor prefix and limit the
2020         -apple prefix to two properties (dashboard-region and line-clamp).
2021
2022         Test: fast/css/limited-vendor-prefix-behavior.html
2023
2024         * css/CSSParser.cpp:
2025         (WebCore::cssPropertyID):
2026
2027 2010-07-21  Adam Barth  <abarth@webkit.org>
2028
2029         Reviewed by Eric Seidel.
2030
2031         Fix the last tree HTML5 tree builder crashes
2032         https://bugs.webkit.org/show_bug.cgi?id=42773
2033
2034         This patch changes the internal representation of a bookmark to handle
2035         the case where one of the adjecent entries in the list of active
2036         formatting elements is actually a marker.
2037
2038         After this patch, the bookmarking mechanism isn't as general, but it
2039         works for the cases we need in the adoption agency.
2040
2041         Also, after this patch, there aren't any more known crashers in the
2042         HTML5 tree builder.  :)
2043
2044         * html/HTMLFormattingElementList.cpp:
2045         (WebCore::HTMLFormattingElementList::bookmarkFor):
2046         (WebCore::HTMLFormattingElementList::swapTo):
2047         * html/HTMLFormattingElementList.h:
2048         (WebCore::HTMLFormattingElementList::Bookmark::Bookmark):
2049         (WebCore::HTMLFormattingElementList::Bookmark::moveToAfter):
2050         (WebCore::HTMLFormattingElementList::Bookmark::hasBeenMoved):
2051         (WebCore::HTMLFormattingElementList::Bookmark::mark):
2052         (WebCore::HTMLFormattingElementList::first):
2053         * html/HTMLTreeBuilder.cpp:
2054         (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):
2055
2056 2010-07-21  Tony Gentilcore  <tonyg@chromium.org>
2057
2058         Unreviewed build fix.
2059
2060         Disable overzealous ASSERT
2061         https://bugs.webkit.org/show_bug.cgi?id=42775
2062
2063         No new tests because no changed functionality.
2064
2065         * loader/FrameLoader.cpp:
2066         (WebCore::FrameLoader::stopLoading):
2067
2068 2010-07-21  Yael Aharon  <yael.aharon@nokia.com>
2069
2070         Reviewed by Darin Adler.
2071
2072         Crash in Notification::disconnectFrame() triggered by Frame::lifeSupportTimerFired()
2073         https://bugs.webkit.org/show_bug.cgi?id=42534
2074
2075         Call NotificationsCenter::disconnectFrame() when the frame is disconnected from the page.
2076         Calling it from the destructor of Frame is too late and sometimes causes access violation.
2077         I was not able to reproduce this crash, so did not add new tests.
2078         This patch is based on the error reported in 
2079         http://code.google.com/p/chromium/issues/detail?id=49323.
2080
2081         * page/DOMWindow.cpp:
2082         (WebCore::DOMWindow::pageDestroyed):
2083         * page/DOMWindow.h:
2084         * page/Frame.cpp:
2085         (WebCore::Frame::pageDestroyed):
2086
2087 2010-07-21  Anders Carlsson  <andersca@apple.com>
2088
2089         Reviewed by Sam Weinig.
2090
2091         Don't assert when clicking on a plug-in in WebKit2
2092         https://bugs.webkit.org/show_bug.cgi?id=42762
2093
2094         Remove a now invalid assert and return false instead.
2095
2096         * page/mac/EventHandlerMac.mm:
2097         (WebCore::EventHandler::passMouseDownEventToWidget):
2098
2099 2010-07-21  Anton Muhin  <antonm@chromium.org>
2100
2101         Reviewed by Adam Barth.
2102
2103         [v8] Revert r60670 as it introduced a regression: in some cases named children couldn't be retrieved.
2104         https://bugs.webkit.org/show_bug.cgi?id=42766
2105
2106         See http://code.google.com/p/chromium/issues/detail?id=48804 for more details.
2107
2108         * bindings/scripts/CodeGeneratorV8.pm:
2109         * bindings/v8/ScriptController.cpp:
2110         (WebCore::ScriptController::namedItemAdded):
2111         (WebCore::ScriptController::namedItemRemoved):
2112         * bindings/v8/V8DOMWindowShell.cpp:
2113         (WebCore::V8DOMWindowShell::updateDocumentWrapperCache):
2114         * bindings/v8/V8DOMWindowShell.h:
2115         * bindings/v8/V8DOMWrapper.cpp:
2116         (WebCore::V8DOMWrapper::instantiateV8Object):
2117         * bindings/v8/custom/V8HTMLDocumentCustom.cpp:
2118         (WebCore::V8HTMLDocument::namedPropertyDeleter):
2119         (WebCore::V8HTMLDocument::namedPropertyGetter):
2120         (WebCore::V8HTMLDocument::indexedPropertyGetter):
2121         (WebCore::V8HTMLDocument::allAccessorSetter):
2122         (WebCore::toV8):
2123
2124 2010-07-21  Eric Carlson  <eric.carlson@apple.com>
2125
2126         Reviewed by Simon Fraser.
2127
2128         Update plug-in proxy backed <video> and <audio> elements.
2129         https://bugs.webkit.org/show_bug.cgi?id=42770
2130         <rdar://problem/7963467>
2131
2132         * html/HTMLMediaElement.cpp:
2133         (WebCore::HTMLMediaElement::attributeChanged): Only call m_player->setControls when
2134         the 'controls' attribute changes.
2135         (WebCore::HTMLMediaElement::createRenderer): 
2136         (WebCore::HTMLMediaElement::attach): If we have a proxy widget, call the WebFrameLoaderClient's 
2137         'hide' and 'show' methods for the media proxy plug-in to make sure we remove and re-add the 
2138         plug-in's view to the global list of plugin views.
2139         (WebCore::HTMLMediaElement::createMediaPlayerProxy): Bail if m_proxyWidget is non-NULL
2140
2141         * loader/EmptyClients.h:
2142         (WebCore::EmptyFrameLoaderClient::hideMediaPlayerProxyPlugin): Empty stub method.
2143         (WebCore::EmptyFrameLoaderClient::showMediaPlayerProxyPlugin): Ditto.
2144
2145         * loader/FrameLoader.cpp:
2146         * loader/FrameLoaderClient.h:
2147         * loader/SubframeLoader.cpp:
2148         (WebCore::FrameLoader::loadMediaPlayerProxyPlugin): Always set m_containsPlugIns in loadMediaPlayerProxyPlugin(),
2149          even when we don't have a renderer.
2150         (WebCore::FrameLoader::hideMediaPlayerProxyPlugin): New, call hideMediaPlayerProxyPlugin.
2151         (WebCore::FrameLoader::showMediaPlayerProxyPlugin): New, call showMediaPlayerProxyPlugin.
2152
2153         * platform/graphics/MediaPlayer.cpp:
2154         (WebCore::NullMediaPlayerPrivate::setControls):
2155         (WebCore::MediaPlayer::setControls): New, pass through to media engine.
2156         (WebCore::MediaPlayer::enterFullscreen): Ditto.
2157         (WebCore::MediaPlayer::exitFullscreen): Ditto.
2158         * platform/graphics/MediaPlayer.h:
2159         * platform/graphics/MediaPlayerPrivate.h:
2160         (WebCore::MediaPlayerPrivateInterface::setControls):
2161         (WebCore::MediaPlayerPrivateInterface::enterFullscreen):
2162         (WebCore::MediaPlayerPrivateInterface::exitFullscreen):
2163
2164 2010-07-21  Dimitri Glazkov  <dglazkov@chromium.org>
2165
2166         Reviewed by Darin Adler.
2167
2168         Make more members of CSSStyleSelector private.
2169         https://bugs.webkit.org/show_bug.cgi?id=42757
2170
2171         No change in behavior, so no new tests.
2172
2173         * css/CSSStyleSelector.h: Made a publicly-unused methods private.
2174
2175 2010-07-21  Pavel Feldman  <pfeldman@chromium.org>
2176
2177         Reviewed by Timothy Hatcher.
2178
2179         Web Inspector: consider jQuery object to be of array nature.
2180
2181         https://bugs.webkit.org/show_bug.cgi?id=42758
2182
2183         * inspector/front-end/InjectedScript.js:
2184         (injectedScriptConstructor):
2185
2186 2010-07-21  Alexey Proskuryakov  <ap@apple.com>
2187
2188         Unreviewed Windows build fix.
2189
2190         https://bugs.webkit.org/show_bug.cgi?id=42717
2191         <rdar://problem/7062824> A wrong password entered for site or proxy auth remains in WebCore
2192         credential storage, and is sent with subsequent requests
2193
2194         * platform/network/cf/ResourceHandleCFNet.cpp:
2195         (WebCore::ResourceHandle::didReceiveAuthenticationChallenge): Don't use that direct a
2196         copy/paste, oops!
2197
2198 2010-07-21  Simon Fraser  <simon.fraser@apple.com>
2199
2200         Reviewed by Darin Adler.
2201
2202         When GraphicsLayer::contentsOrientation() is CompositingCoordinatesBottomUp, need to flip repaint rects
2203         https://bugs.webkit.org/show_bug.cgi?id=42662
2204         
2205         If the contentsOrientation() on a GraphicsLayer is CompositingCoordinatesBottomUp, then we
2206         need to flip the rects passed to setNeedsDisplayInRect:.
2207         
2208         Avoid writing this code twice by making a bare function to share code
2209         between WebLayer and WebTiledLayer. Convert the existing +drawContents:ofLayer:intoContext:
2210         class method into a bare function, because the Obj-C calling overhead doesn't buy us anything.
2211         
2212         Take out an assertion in GraphicsLayerCA::updateContentsTransform() that is not
2213         correct.
2214         
2215         No new tests because no layers use bottom-up contents at present.
2216
2217         * platform/graphics/mac/GraphicsLayerCA.mm:
2218         (WebCore::GraphicsLayerCA::updateContentsTransform):
2219         * platform/graphics/mac/WebLayer.h:
2220         * platform/graphics/mac/WebLayer.mm:
2221         (drawLayerContents):
2222         (setLayerNeedsDisplayInRect):
2223         (-[WebLayer setNeedsDisplayInRect:]):
2224         (-[WebLayer drawInContext:]):
2225         * platform/graphics/mac/WebTiledLayer.mm:
2226         (-[WebTiledLayer setNeedsDisplayInRect:]):
2227         (-[WebTiledLayer drawInContext:]):
2228
2229 2010-07-20  Alexey Proskuryakov  <ap@apple.com>
2230
2231         Reviewed by Darin Adler.
2232
2233         https://bugs.webkit.org/show_bug.cgi?id=42717
2234         <rdar://problem/7062824> A wrong password entered for site or proxy auth remains in WebCore
2235         credential storage, and is sent with subsequent requests
2236
2237         Tests: http/tests/security/401-logout/401-logout.php
2238                http/tests/xmlhttprequest/remember-bad-password.html
2239
2240         * platform/network/CredentialStorage.cpp: (WebCore::CredentialStorage::remove):
2241         * platform/network/CredentialStorage.h:
2242         Added a way to remove stored credentials for a given protection space.
2243
2244         * platform/network/cf/ResourceHandleCFNet.cpp: (WebCore::ResourceHandle::didReceiveAuthenticationChallenge):
2245         * platform/network/mac/ResourceHandleMac.mm: (WebCore::ResourceHandle::didReceiveAuthenticationChallenge):
2246         Remove stored credentials if they didn't work the first time.
2247
2248 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2249
2250         Reviewed by Adam Roben.
2251
2252         WebKit on Windows should build optionally with an unversioned ICU DLL
2253         https://bugs.webkit.org/show_bug.cgi?id=42722
2254         <rdar://problem/8211767> WebKit needs to link against unversioned ICU
2255
2256         To get the proper value for U_DISABLE_RENAMING into all source files, we force
2257         the include of ICUVersion.h (our generated header) via the compiler options.
2258         
2259         * WebCore.vcproj/WebCore.vcproj:
2260         Add forced include of ICUVersion.h.
2261         * WebCore.vcproj/WebCoreCommon.vsprops: Add forced include of ICUVersion.h.
2262
2263 2010-07-21  Ilya Tikhonovsky  <loislo@chromium.org>
2264
2265         Reviewed by Pavel Feldman.
2266
2267         WebInspector: Serialization to JSON in InspectorValue works incorrect
2268         if comma is assigned as decimal separator. Windows with Russian locale.
2269         https://bugs.webkit.org/show_bug.cgi?id=42755
2270
2271         * inspector/InspectorValues.cpp:
2272         (WebCore::InspectorBasicValue::writeJSON):
2273
2274 2010-07-20  Simon Fraser  <simon.fraser@apple.com>
2275
2276         Reviewed by Dan Bernstein.
2277
2278         CSS rotation transform can cause elements with certain styles to vanish during rotation.
2279         https://bugs.webkit.org/show_bug.cgi?id=42579
2280         
2281         Remove some untested, broken code that, as far as I can tell, could never get hit.
2282         Use the correct rect for the PaintInfo so that zero-sized elements with overflow:hidden
2283         but a border paint correctly.
2284
2285         Test: compositing/overflow/zero-size-overflow.html
2286
2287         * rendering/RenderLayerBacking.cpp:
2288         (WebCore::RenderLayerBacking::paintIntoLayer):
2289
2290 2010-07-21  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2291
2292         Reviewed by Laszlo Gombos.
2293
2294         [QT] Leak and few more fixes to qt port of geolocation
2295         https://bugs.webkit.org/show_bug.cgi?id=42753
2296
2297         Fixes memory leak and default values set to attributes. 
2298         
2299         Memory leak fix no new test case added. 
2300
2301         * platform/qt/GeolocationServiceQt.cpp: Added.
2302         (WebCore::GeolocationServiceQt::~GeolocationServiceQt):
2303         (WebCore::GeolocationServiceQt::positionUpdated):
2304
2305 2010-07-21  Kristian Amlie <kristian.amlie@nokia.com>
2306
2307         Reviewed by Kenneth Rohde Christiansen.
2308
2309         Added automatic sqlite extraction for Symbian to QtWebKit.
2310
2311         Also added sqlite detection in case sqlite is not present in the SDK.
2312         This is possible if WebKit is compiled standalone.
2313
2314         The inclusion part is a consequence of commit c578c6c1d6d in the Qt
2315         repository. It will not work on Qt versions < 4.7.1, but that is ok,
2316         since the only build system it will affect is marked as experimental
2317         in the whole 4.7 series.
2318
2319         * WebCore.pro:
2320
2321 2010-07-21  Adam Barth  <abarth@webkit.org>
2322
2323         Reviewed by Eric Seidel.
2324
2325         Associate elements with the active form
2326         https://bugs.webkit.org/show_bug.cgi?id=42728
2327
2328         This patch fixes fast/forms/formmove3.html.  The test still doesn't
2329         pass due to some render tree differences, but it works as intended now.
2330
2331         To fix this test, I needed to deviate from the spec slight.  Minefield
2332         seems to have the same deviation:
2333           - http://www.w3.org/Bugs/Public/show_bug.cgi?id=10216
2334
2335         * html/HTMLConstructionSite.cpp:
2336         (WebCore::HTMLConstructionSite::takeForm):
2337         (WebCore::HTMLConstructionSite::setForm):
2338         (WebCore::HTMLConstructionSite::createHTMLElement):
2339         * html/HTMLConstructionSite.h:
2340         (WebCore::HTMLConstructionSite::form):
2341         * html/HTMLTreeBuilder.cpp:
2342         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
2343
2344 2010-07-21  Adam Barth  <abarth@webkit.org>
2345
2346         Reviewed by Eric Seidel.
2347
2348         The adoption agency doesn't properly attach()
2349         https://bugs.webkit.org/show_bug.cgi?id=42727
2350
2351         The adoption agency is transliterated rather directly from the spec,
2352         but it misses some of the WebKit-specific machinations, such as
2353         attaching to the render tree.
2354
2355         The algorithm, as written, is a minor layer violation.  I've added to
2356         the problem by calling attach() from HTMLTreeBuilder (even though
2357         that's the job of the HTMLConstructionSite).  We'll need to clean up
2358         the layering a bit at some point.
2359
2360         This patch half fixes fast/forms/formmove3.html.  Hopefully I'll finish
2361         fixing it in the next patch.
2362
2363         * html/HTMLConstructionSite.cpp:
2364         (WebCore::HTMLConstructionSite::createHTMLElementFromSavedElement):
2365         * html/HTMLTreeBuilder.cpp:
2366         (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):
2367
2368 2010-07-21  Adam Barth  <abarth@webkit.org>
2369
2370         Reviewed by Eric Seidel.
2371
2372         <input> elements with no type attribute should be foster parented
2373         https://bugs.webkit.org/show_bug.cgi?id=42725
2374
2375         Fix spec transcription error.
2376
2377         * html/HTMLTreeBuilder.cpp:
2378         (WebCore::HTMLTreeBuilder::processStartTagForInTable):
2379
2380 2010-07-21  Adam Barth  <abarth@webkit.org>
2381
2382         Reviewed by Eric Seidel.
2383
2384         Fix fast/css/last-child-style-sharing.html
2385         https://bugs.webkit.org/show_bug.cgi?id=42731
2386
2387         Prior to this patch, we weren't calling finishParsingChildren on the
2388         body element.  We need a more systematic way of catching these bugs.
2389
2390         * html/HTMLElementStack.cpp:
2391         (WebCore::HTMLElementStack::popAll):
2392         * html/HTMLElementStack.h:
2393         * html/HTMLTreeBuilder.cpp:
2394         (WebCore::HTMLTreeBuilder::processEndOfFile):
2395
2396 2010-07-21  Hans Wennborg  <hans@chromium.org>
2397
2398         Reviewed by Steve Block.
2399
2400         Runtime feature switch for device orientation
2401         https://bugs.webkit.org/show_bug.cgi?id=42265
2402
2403         Add a runtime feature switch that decides whether device orientation
2404         events are available or not. Defaults to true.
2405
2406         * WebCore.vcproj/WebCoreCommon.vsprops: Add bindings/generic to include path.
2407         * bindings/generic/RuntimeEnabledFeatures.cpp:
2408         * bindings/generic/RuntimeEnabledFeatures.h:
2409         (WebCore::RuntimeEnabledFeatures::setDeviceOrientationEnabled):
2410         (WebCore::RuntimeEnabledFeatures::deviceOrientationEnabled):
2411         (WebCore::RuntimeEnabledFeatures::deviceOrientationEventEnabled):
2412         (WebCore::RuntimeEnabledFeatures::ondeviceorientationEnabled):
2413         * page/DOMWindow.cpp:
2414         (WebCore::DOMWindow::addEventListener):
2415         (WebCore::DOMWindow::removeEventListener):
2416         (WebCore::DOMWindow::removeAllEventListeners):
2417         * page/DOMWindow.idl:
2418         * page/Page.cpp:
2419         (WebCore::Page::Page):
2420
2421 2010-07-21  Zoltan Herczeg  <zherczeg@webkit.org>
2422
2423         Reviewed Nikolas Zimmermann.
2424
2425         SVGFilterElement & SVGFE*Element don't support dynamic invalidation, when attributes change
2426         https://bugs.webkit.org/show_bug.cgi?id=42244
2427
2428         The implementation was done by Nikolas Zimmermann before, but
2429         there was no tests for it. The patch also implements the simple
2430         setFilterRes method.
2431
2432         Tests: svg/dynamic-updates/SVGFilterElement-dom-filterRes-attr.html
2433                svg/dynamic-updates/SVGFilterElement-dom-filterUnits-attr.html
2434                svg/dynamic-updates/SVGFilterElement-dom-height-attr.html
2435                svg/dynamic-updates/SVGFilterElement-dom-primitiveUnits-attr.html
2436                svg/dynamic-updates/SVGFilterElement-dom-width-attr.html
2437                svg/dynamic-updates/SVGFilterElement-dom-x-attr.html
2438                svg/dynamic-updates/SVGFilterElement-dom-y-attr.html
2439                svg/dynamic-updates/SVGFilterElement-svgdom-filterRes-call.html
2440                svg/dynamic-updates/SVGFilterElement-svgdom-filterResX-prop.html
2441                svg/dynamic-updates/SVGFilterElement-svgdom-filterResY-prop.html
2442                svg/dynamic-updates/SVGFilterElement-svgdom-filterUnits-prop.html
2443                svg/dynamic-updates/SVGFilterElement-svgdom-height-prop.html
2444                svg/dynamic-updates/SVGFilterElement-svgdom-primitiveUnits-prop.html
2445                svg/dynamic-updates/SVGFilterElement-svgdom-width-prop.html
2446                svg/dynamic-updates/SVGFilterElement-svgdom-x-prop.html
2447                svg/dynamic-updates/SVGFilterElement-svgdom-y-prop.html
2448
2449         * svg/SVGFilterElement.cpp:
2450         (WebCore::SVGFilterElement::setFilterRes):
2451         (WebCore::SVGFilterElement::svgAttributeChanged):
2452         (WebCore::SVGFilterElement::childrenChanged):
2453         * svg/SVGFilterElement.h:
2454
2455 2010-07-21  Ben Murdoch  <benm@google.com>
2456
2457         Unreviewed, build fix.
2458
2459         Forgot to update the role of HitTestRequest.h in
2460         the xcode project after landing patch from
2461         https://bugs.webkit.org/show_bug.cgi?id=39493
2462
2463         * WebCore.xcodeproj/project.pbxproj: Make HitTestRequest.h
2464             private so it can be included in WebKit by EventHandler.h
2465             which is already private..
2466
2467 2010-07-20  Ben Murdoch  <benm@google.com>
2468
2469         Reviewed by Steve Block.
2470
2471         Touch events do not affect the :active CSS state
2472         https://bugs.webkit.org/show_bug.cgi?id=39493
2473
2474         Test: fast/events/touch/touch-active-state.html
2475
2476         * WebCore.exp.in: Update exports for new signature of
2477             hitTestResultAtPoint.
2478         * WebCore.order: ditto.
2479         * page/EventHandler.cpp:
2480         (WebCore::EventHandler::EventHandler):
2481         (WebCore::EventHandler::hitTestResultAtPoint): Pass the type
2482             of the hit test to perform as a parameter with a default
2483             value rather than harcoding it in the function body.
2484         (WebCore::EventHandler::handleMouseMoveEvent): Do not modiify
2485             the active element during a mouse move if the user is
2486             touching the screen.
2487         (WebCore::EventHandler::handleTouchEvent): Set the correct
2488             type of hit test to perform depending on the type of the
2489             touch event we are handling.
2490         * page/EventHandler.h: Update the signature of hitTestResultAtPoint.
2491
2492 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
2493
2494         Reviewed by Pavel Feldman.
2495
2496         console.assert should include stack trace with line numbers.
2497         https://bugs.webkit.org/show_bug.cgi?id=22556
2498
2499         Test: inspector/console-assert.html
2500
2501         * bindings/v8/ScriptCallStack.cpp:
2502         (WebCore::getTopFrameLocation):
2503         (WebCore::toScriptCallFramesVector):
2504         (WebCore::ScriptCallStack::create):
2505         (WebCore::ScriptCallStack::ScriptCallStack):
2506         * bindings/v8/ScriptCallStack.h:
2507         * bindings/v8/ScriptController.cpp:
2508         (WebCore::ScriptController::setCaptureCallStackForUncaughtExceptions):
2509         * bindings/v8/custom/V8ConsoleCustom.cpp:
2510         (WebCore::V8Console::traceCallback):
2511         (WebCore::V8Console::assertCallback):
2512         * inspector/ConsoleMessage.cpp:
2513         (WebCore::ConsoleMessage::ConsoleMessage):
2514         * inspector/InspectorController.cpp:
2515         (WebCore::InspectorController::addMessageToConsole):
2516         * inspector/front-end/ConsoleView.js:
2517         (WebInspector.ConsoleMessage.prototype._formatMessage):
2518         (WebInspector.ConsoleMessage.prototype.toMessageElement):
2519         * page/Console.cpp:
2520         (WebCore::Console::lastWMLErrorMessage):
2521         * page/Console.idl:
2522
2523 2010-07-20  Rafael Antognolli  <antognolli@profusion.mobi>
2524
2525         Reviewed by Kent Tamura.
2526
2527         [EFL] Use log functions instead of fprintf
2528         https://bugs.webkit.org/show_bug.cgi?id=42576
2529
2530         Use LOG and LOG_ERROR instead of fprintf.
2531
2532         No new tests, no new functionality.
2533
2534         * platform/efl/WidgetEfl.cpp:
2535         (WebCore::Widget::applyFallbackCursor):
2536
2537 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
2538
2539         Reviewed by Yury Semikhatsky.
2540
2541         WebInspector: the bindings generation helper script can be
2542         simplified a bit. Chromium.
2543         https://bugs.webkit.org/show_bug.cgi?id=42523
2544
2545         * WebCore.gyp/WebCore.gyp:
2546         * WebCore.gyp/scripts/rule_binding.py:
2547         * bindings/scripts/CodeGenerator.pm:
2548         * bindings/scripts/CodeGeneratorCPP.pm:
2549         * bindings/scripts/CodeGeneratorJS.pm:
2550         * bindings/scripts/CodeGeneratorObjC.pm:
2551         * bindings/scripts/CodeGeneratorV8.pm:
2552         * bindings/scripts/generate-bindings.pl:
2553         * inspector/CodeGeneratorInspector.pm:
2554
2555 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2556
2557         Re-save vsprops file after no-op edit in Visual Studio
2558         to fix manual edit issues.
2559
2560         * WebCore.vcproj/WebCoreCommon.vsprops:
2561
2562 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2563
2564         Re-save vcproj file after no-op edit in Visual Studio
2565         to fix manual edit issues.
2566
2567         * WebCore.vcproj/WebCore.vcproj:
2568
2569 2010-07-20  Adam Barth  <abarth@webkit.org>
2570
2571         Reviewed by Eric Seidel.
2572
2573         Include attributes when reconstructing elements in HTML5 tree builder
2574         https://bugs.webkit.org/show_bug.cgi?id=42594
2575
2576         Remove FIXME now that we have a test.
2577
2578         * html/HTMLConstructionSite.cpp:
2579         (WebCore::HTMLConstructionSite::createHTMLElementFromElementRecord):
2580
2581 2010-07-20  Sam Weinig  <sam@webkit.org>
2582
2583         Reviewed by Brady Eidson.
2584
2585         Patch for https://bugs.webkit.org/show_bug.cgi?id=42719
2586         Make Acid2 pass in WebKit2
2587
2588         * WebCore.exp.in: Update exported functions.
2589
2590 2010-07-20  Gavin Peters  <gavinp@chromium.org>
2591
2592         Reviewed by Adam Barth.
2593
2594         prefetch categorization is exactly wrong
2595         https://bugs.webkit.org/show_bug.cgi?id=42651
2596
2597         While I was in the area, I saw that CachedResource::schedule() was dead, so I have removed it.
2598
2599         Test: fast/dom/HTMLLinkElement/onload-completion-test.html
2600
2601         * loader/CachedCSSStyleSheet.h:
2602         * loader/CachedFont.h:
2603         * loader/CachedImage.h:
2604         * loader/CachedResource.cpp:
2605         * loader/CachedResource.h:
2606         (WebCore::CachedResource::isPrefetch):
2607         * loader/CachedScript.h:
2608         * loader/CachedXBLDocument.h:
2609         * loader/CachedXSLStyleSheet.h:
2610
2611 2010-07-20  James Hawkins  <jhawkins@chromium.org>
2612
2613         Reviewed by Darin Fisher.
2614
2615         Expose the form submission trigger on the HTMLFormElement object. This
2616         is used to verify that the user submitted the form instead of JS when
2617         saving form data in Chrome AutoFill.
2618         https://bugs.webkit.org/show_bug.cgi?id=42479
2619
2620         No new tests as this is only used by the Chromium WebKit API.
2621
2622         * html/HTMLFormElement.cpp:
2623         (WebCore::HTMLFormElement::HTMLFormElement):
2624         (WebCore::HTMLFormElement::submit):
2625         (WebCore::HTMLFormElement::reset):
2626         (WebCore::HTMLFormElement::submissionTrigger):
2627         * html/HTMLFormElement.h:
2628
2629 2010-07-20  Adam Barth  <abarth@webkit.org>
2630
2631         Unreviewed.
2632
2633         Fix fast/dom/title-content-write-set.html for HTML5 tree builder
2634         https://bugs.webkit.org/show_bug.cgi?id=42668
2635
2636         Address some late-breaking review comments.
2637
2638         * dom/CharacterData.cpp:
2639         (WebCore::CharacterData::appendDataCommon):
2640         (WebCore::CharacterData::parserAppendData):
2641         (WebCore::CharacterData::appendData):
2642         (WebCore::CharacterData::insertData):
2643         (WebCore::CharacterData::replaceData):
2644         * dom/CharacterData.h:
2645
2646 2010-07-20  Daniel Erat  <derat@chromium.org>
2647
2648         Reviewed by Ojan Vafai.
2649
2650         Subpixel rendering always disabled for Chromium Linux
2651         https://bugs.webkit.org/show_bug.cgi?id=42220
2652
2653         Explicitly initialize global Chromium Linux isSkiaSubpixelGlyphs
2654         flag to false for clarity.
2655
2656         * platform/graphics/chromium/FontPlatformDataLinux.cpp:
2657
2658 2010-07-20  Adam Barth  <abarth@webkit.org>
2659
2660         Reviewed by Darin Adler.
2661
2662         Fix fast/dom/title-content-write-set.html for HTML5 tree builder
2663         https://bugs.webkit.org/show_bug.cgi?id=42668
2664
2665         When I split parserAppendData from appendData, it was tempting to cut
2666         the function at dispatchModifiedEvent, but that's not quite right.  We
2667         still need to notify the parent that it's children have changed.
2668
2669         * dom/CharacterData.cpp:
2670         (WebCore::CharacterData::appendDataCommon):
2671         (WebCore::CharacterData::parserAppendData):
2672         (WebCore::CharacterData::appendData):
2673         * dom/CharacterData.h:
2674
2675 2010-07-20  Chris Fleizach  <cfleizach@apple.com>
2676
2677         Reviewed by Beth Dakin.
2678
2679         CrashTracer: [USER] 300 crashes in Safari at com.apple.WebCore: WebCore::AccessibilityTable::isTableExposableThroughAccessibility + 573
2680         https://bugs.webkit.org/show_bug.cgi?id=42652
2681
2682         When a table cell accesses its parent table, we should not use getOrCreate, because creating an AXTable inspects its render tree state
2683         which may be out of date, leading to a crash.
2684         By using only get(), it implies that the AXTable must be created before AXTableCells. This should
2685         always be the case when AT clients access a table.
2686
2687         Test: accessibility/updating-attribute-in-table-causes-crash.html
2688
2689         * accessibility/AccessibilityTableCell.cpp:
2690         (WebCore::AccessibilityTableCell::parentTable):
2691
2692 2010-07-20  Abhishek Arya  <inferno@chromium.org>
2693
2694         Reviewed by David Hyatt.
2695
2696         Check the node is a text node before doing the static cast
2697         for editing commands.
2698         https://bugs.webkit.org/show_bug.cgi?id=42655
2699
2700         Test: editing/execCommand/editing-nontext-node-crash.xhtml
2701
2702         * editing/DeleteSelectionCommand.cpp:
2703         (WebCore::DeleteSelectionCommand::fixupWhitespace):
2704         * editing/InsertLineBreakCommand.cpp:
2705         (WebCore::InsertLineBreakCommand::doApply):
2706         * editing/InsertParagraphSeparatorCommand.cpp:
2707         (WebCore::InsertParagraphSeparatorCommand::doApply):
2708
2709 2010-07-20  Leo Yang  <leo.yang@torchmobile.com.cn>
2710
2711         Reviewed by David Hyatt.
2712
2713         Don't merge Anonymous block whose first child is inline run-in.
2714         Make run-in recalculate its style after its renderer is destroyed.
2715         https://bugs.webkit.org/show_bug.cgi?id=41375.
2716
2717         Test: fast/runin/crash-when-reparent-sibling.html
2718
2719         * rendering/RenderBlock.cpp:
2720         (WebCore::canMergeContiguousAnonymousBlocks):
2721         * rendering/RenderObjectChildList.cpp:
2722         (WebCore::RenderObjectChildList::destroyLeftoverChildren):
2723
2724 2010-07-20  Steve Block  <steveblock@google.com>
2725
2726         Unreviewed Qt test fix.
2727
2728         Qt should not use PREEMPT_GEOLOCATION_PERMISSION until Bug 42027 is fixed.
2729         https://bugs.webkit.org/show_bug.cgi?id=42068
2730
2731         No new tests, test fix only.
2732
2733         * WebCore.pro:
2734
2735 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
2736
2737         Unreviewed, rolling out r63764.
2738         http://trac.webkit.org/changeset/63764
2739         https://bugs.webkit.org/show_bug.cgi?id=42658
2740
2741         have broken Chromium linux build (Requested by loislo on
2742         #webkit).
2743
2744         * WebCore.gyp/WebCore.gyp:
2745         * WebCore.gyp/scripts/rule_binding.py:
2746         * bindings/scripts/CodeGenerator.pm:
2747         * bindings/scripts/CodeGeneratorCPP.pm:
2748         * bindings/scripts/CodeGeneratorJS.pm:
2749         * bindings/scripts/CodeGeneratorObjC.pm:
2750         * bindings/scripts/CodeGeneratorV8.pm:
2751         * bindings/scripts/generate-bindings.pl:
2752         * inspector/CodeGeneratorInspector.pm:
2753
2754 2010-07-20  Alexey Proskuryakov  <ap@apple.com>
2755
2756         Reviewed by Brady Eidson.
2757
2758         https://bugs.webkit.org/show_bug.cgi?id=41531
2759         Asynchronous cross origin XMLHttpRequest doesn't expose 401 response when withCredentials is false
2760
2761         This doesn't match Firefox, but it matches our sync case, XHR2 spec and common sense.
2762
2763         Test: http/tests/xmlhttprequest/cross-origin-no-authorization.html (updated results).
2764
2765         * loader/DocumentThreadableLoader.cpp:
2766         (WebCore::DocumentThreadableLoader::didReceiveAuthenticationChallenge): Instead of canceling
2767         the request, continue without credentials - if the platform has a necessary method on
2768         ResourceHandle.
2769
2770         * loader/SubresourceLoader.cpp:
2771         (WebCore::SubresourceLoader::didReceiveAuthenticationChallenge): Don't ask resource loader
2772         client for credentials if subresource loader client already took care of those.
2773
2774         * platform/network/ResourceHandle.cpp: (WebCore::ResourceHandle::hasAuthenticationChallenge):
2775         * platform/network/ResourceHandle.h:
2776         Added an accessor to check whether ResourceHandle is currently waiting for credentials.
2777
2778 2010-07-20  Leandro Pereira  <leandro@profusion.mobi>
2779
2780         [EFL] Unreviewed build fix.
2781
2782         Build SVGPathBuilder, SVGPathParser and SVGPathSegListBuilder.
2783         Add websockets include directory.
2784
2785         The EFL port doesn't yet support automated tests.
2786
2787         * CMakeLists.txt:
2788
2789 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
2790
2791         Reviewed by Yury Semikhatsky.
2792
2793         WebInspector: the bindings generation helper script can be
2794         simplified a bit. Chromium.
2795         https://bugs.webkit.org/show_bug.cgi?id=42523
2796
2797         * WebCore.gyp/WebCore.gyp:
2798         * WebCore.gyp/scripts/rule_binding.py:
2799         * bindings/scripts/CodeGenerator.pm:
2800         * bindings/scripts/CodeGeneratorCPP.pm:
2801         * bindings/scripts/CodeGeneratorJS.pm:
2802         * bindings/scripts/CodeGeneratorObjC.pm:
2803         * bindings/scripts/CodeGeneratorV8.pm:
2804         * bindings/scripts/generate-bindings.pl:
2805         * inspector/CodeGeneratorInspector.pm:
2806
2807 2010-07-20  Joseph Pecoraro  <joepeck@webkit.org>
2808
2809         Reviewed by Geoffrey Garen.
2810
2811         WebScriptObject Should Allow Safely Checking For Key Existence
2812         https://bugs.webkit.org/show_bug.cgi?id=42613
2813
2814         Test: platform/mac/fast/objc/webScriptObject-hasWebScriptKey.html
2815
2816         Add private API "hasWebScriptKey" to check for key existence in
2817         a WebScriptObject. Like JavaScript's `in` syntax. This is intended
2818         to be made public eventually.
2819
2820         * bindings/objc/WebScriptObject.mm:
2821         (-[WebScriptObject hasWebScriptKey:]):
2822         * bindings/objc/WebScriptObjectPrivate.h:
2823
2824 2010-07-20  Adam Barth  <abarth@webkit.org>
2825
2826         Reviewed by Eric Seidel.
2827
2828         HTML5 tree builder needs to call dispatchDocumentElementAvailable
2829         https://bugs.webkit.org/show_bug.cgi?id=42654
2830
2831         This patch fixes the follout LayoutTests with --html5-treebuilder:
2832           - userscripts/script-not-run-for-fragments.html
2833           - userscripts/script-run-at-start.html
2834
2835         * html/HTMLConstructionSite.cpp:
2836         (WebCore::HTMLConstructionSite::HTMLConstructionSite):
2837         (WebCore::HTMLConstructionSite::dispatchDocumentElementAvailableIfNeeded):
2838         (WebCore::HTMLConstructionSite::insertHTMLHtmlStartTagBeforeHTML):
2839         (WebCore::HTMLConstructionSite::insertHTMLHtmlElement):
2840         * html/HTMLConstructionSite.h:
2841         * html/HTMLTreeBuilder.cpp:
2842         (WebCore::HTMLTreeBuilder::HTMLTreeBuilder):
2843
2844 2010-07-20  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2845
2846         Reviewed by Steve Block.
2847
2848         Need to be able to configure Geolocation policy regarding user permissions
2849         https://bugs.webkit.org/show_bug.cgi?id=42068
2850
2851         Introducing new USE() flag PREEMPT_GEOLOCATION_PREMISSION using which
2852         acquires user permission first before starting location service
2853
2854         This change does not introduce any change in behavior for any platform. So there are no new tests
2855         added.
2856
2857         * WebCore.pro:
2858         * page/Geolocation.cpp:
2859         (WebCore::Geolocation::startRequest):
2860         (WebCore::Geolocation::setIsAllowed):
2861         (WebCore::Geolocation::startUpdating):
2862         * page/Geolocation.h:
2863
2864 2010-07-20  Adam Barth  <abarth@webkit.org>
2865
2866         Reviewed by Darin Adler.
2867
2868         HTML5 tree builder should restore form state
2869         https://bugs.webkit.org/show_bug.cgi?id=42644
2870
2871         We need to tell self-closing tags that we're done parsing their
2872         children.  This patch fixes the following LayoutTests when run with
2873         --html5-treebuilder:
2874
2875           fast/forms/button-state-restore.html
2876           fast/forms/state-restore-to-non-autocomplete-form.html
2877           fast/forms/state-restore-to-non-edited-controls.html
2878           fast/history/saves-state-after-fragment-nav.html
2879           http/tests/navigation/restore-form-state-https.html
2880
2881         * html/HTMLConstructionSite.cpp:
2882         (WebCore::HTMLConstructionSite::insertSelfClosingHTMLElement):
2883
2884 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
2885
2886         Unreviewed, rolling out r63750.
2887         http://trac.webkit.org/changeset/63750
2888         https://bugs.webkit.org/show_bug.cgi?id=42648
2889
2890         This revision breaks the windows builds (Requested by
2891         hwennborg on #webkit).
2892
2893         * bindings/generic/RuntimeEnabledFeatures.cpp:
2894         * bindings/generic/RuntimeEnabledFeatures.h:
2895         * page/DOMWindow.cpp:
2896         (WebCore::DOMWindow::addEventListener):
2897         (WebCore::DOMWindow::removeEventListener):
2898         (WebCore::DOMWindow::removeAllEventListeners):
2899         * page/DOMWindow.idl:
2900         * page/Page.cpp:
2901         (WebCore::Page::Page):
2902
2903 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
2904
2905         Reviewed by Pavel Feldman.
2906
2907         WebInspector: It is possible to show full call stack instead of top frame for Caller
2908         and Call Site properties in Timeline panel.
2909         https://bugs.webkit.org/show_bug.cgi?id=42620
2910
2911         * English.lproj/localizedStrings.js:
2912         * inspector/front-end/TimelinePanel.js:
2913         (WebInspector.TimelinePanel.FormattedRecord):
2914         (WebInspector.TimelinePanel.FormattedRecord.prototype._generatePopupContent):
2915         (WebInspector.TimelinePanel.FormattedRecord.prototype._getRecordDetails):
2916         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendElementRow):
2917         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendLinkRow):
2918         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendStackTrace):
2919         * inspector/front-end/inspector.css:
2920         (.timeline-details):
2921         (.timeline-function-name):
2922         (.timeline-stacktrace-title):
2923
2924 2010-07-20  Leon Clarke  <leonclarke@google.com>
2925
2926         Reviewed by Pavel Feldman.
2927
2928         Make things compile again when the inspector is disabled, following
2929         recent inspector improvements.
2930         https://bugs.webkit.org/show_bug.cgi?id=42632
2931
2932         No new tests. Fixing a build break.
2933
2934
2935         * bindings/v8/ScriptCallStack.cpp:
2936         (WebCore::ScriptCallStack::stackTrace):
2937         * inspector/InspectorController.h:
2938         (WebCore::InspectorController::didInsertDOMNode):
2939         (WebCore::InspectorController::didRemoveDOMNode):
2940         (WebCore::InspectorController::didModifyDOMAttr):
2941         * inspector/InspectorDOMAgent.h:
2942         * loader/appcache/ApplicationCacheGroup.cpp:
2943
2944 2010-07-20  Anton Muhin  <antonm@chromium.org>
2945
2946         Reviewed by Adam Barth.
2947
2948         [v8] Allow handles to be disposed and WebKit objects to be dereferenced even if their map is already destroyed.
2949         https://bugs.webkit.org/show_bug.cgi?id=42634
2950
2951         Currently DOMDataStore could be destroyed even if it has some mappings (it gets destroyed
2952         when its isolated context gets GCed).  However in this case, handles allocated for
2953         such objects would never be disposed as we require presence of mapping from wrapped
2954         WebKit object to handle being collected in the map and now map is gone.  That leads to
2955         zombie objects in both WebKit (wrapped WebKit object doesn't get dereferenced) and V8
2956         (both handle and V8 wrapper object could not be destroyed).
2957
2958         See http://code.google.com/p/chromium/issues/detail?id=47125 for further discussion.
2959
2960         * bindings/v8/DOMData.h:
2961         (WebCore::DOMData::handleWeakObject):
2962         * bindings/v8/DOMDataStore.cpp:
2963         (WebCore::DOMDataStore::weakNodeCallback):
2964         * bindings/v8/V8DOMMap.h:
2965         (WebCore::WeakReferenceMap::~WeakReferenceMap):
2966
2967 2010-07-20  Hans Wennborg  <hans@chromium.org>
2968
2969         Reviewed by Steve Block.
2970
2971         Runtime feature switch for device orientation
2972         https://bugs.webkit.org/show_bug.cgi?id=42265
2973
2974         Add a runtime feature switch that decides whether device orientation
2975         events are available or not. Defaults to true.
2976
2977         * bindings/generic/RuntimeEnabledFeatures.cpp:
2978         * bindings/generic/RuntimeEnabledFeatures.h:
2979         (WebCore::RuntimeEnabledFeatures::setDeviceOrientationEnabled):
2980         (WebCore::RuntimeEnabledFeatures::deviceOrientationEnabled):
2981         (WebCore::RuntimeEnabledFeatures::deviceOrientationEventEnabled):
2982         (WebCore::RuntimeEnabledFeatures::ondeviceorientationEnabled):
2983         * page/DOMWindow.cpp:
2984         (WebCore::DOMWindow::addEventListener):
2985         (WebCore::DOMWindow::removeEventListener):
2986         (WebCore::DOMWindow::removeAllEventListeners):
2987         * page/DOMWindow.idl:
2988         * page/Page.cpp:
2989         (WebCore::Page::Page):
2990
2991 2010-07-20  Hayato Ito  <hayato@chromium.org>
2992
2993         Reviewed by Darin Adler.
2994
2995         Fixed a crash when deeply nested CSS selector is used.
2996         https://bugs.webkit.org/show_bug.cgi?id=41129
2997
2998         This patch deletes CSSSelectors iteratively so that it doesn't cause stack overflow.
2999
3000         Test: fast/css/css-selector-deeply-nested.html
3001
3002         * css/CSSSelector.cpp:
3003         (WebCore::CSSSelectorBag::~CSSSelectorBag):
3004         (WebCore::CSSSelectorBag::isEmpty):
3005         (WebCore::CSSSelectorBag::append):
3006         (WebCore::CSSSelectorBag::takeAny):
3007         (WebCore::CSSSelector::~CSSSelector):
3008         (WebCore::CSSSelector::specificity):
3009         * css/CSSSelector.h:
3010
3011 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
3012
3013         Reviewed by Pavel Feldman.
3014
3015         [v8] Web Inspector: inspected page renderer crashes when inspected page has JS syntax error
3016         https://bugs.webkit.org/show_bug.cgi?id=42642
3017
3018         * bindings/v8/V8ConsoleMessage.cpp:
3019         (WebCore::V8ConsoleMessage::handler):
3020
3021 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
3022
3023         Unreviewed, rolling out r63742.
3024         http://trac.webkit.org/changeset/63742
3025         https://bugs.webkit.org/show_bug.cgi?id=42641
3026
3027         Broke Leopard Intel build. (Requested by bbandix on #webkit).
3028
3029         * WebCore.pro:
3030         * page/Geolocation.cpp:
3031         (WebCore::Geolocation::startRequest):
3032         (WebCore::Geolocation::setIsAllowed):
3033         (WebCore::Geolocation::startUpdating):
3034         * page/Geolocation.h:
3035
3036 2010-07-20  Adam Roben  <aroben@apple.com>
3037
3038         Windows build fix
3039
3040         * WebCore.vcproj/WebCore.vcproj: Fixed case of "Name" attribute.
3041
3042 2010-07-20  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
3043
3044         Reviewed by Steve Block.
3045
3046         Need to be able to configure Geolocation policy regarding user permissions
3047         https://bugs.webkit.org/show_bug.cgi?id=42068
3048
3049         Introducing new USE() flag PREEMPT_GEOLOCATION_PREMISSION using which
3050         acquires user permission first before starting location service
3051
3052         This change does not introduce any change in behavior for any platform. So there are no new tests
3053         added.
3054
3055         * WebCore.pro:
3056         * page/Geolocation.cpp:
3057         (WebCore::Geolocation::startRequest):
3058         (WebCore::Geolocation::setIsAllowed):
3059         (WebCore::Geolocation::startUpdating):
3060         * page/Geolocation.h:
3061
3062 2010-07-20  Alexander Pavlov  <apavlov@chromium.org>
3063
3064         Reviewed by Pavel Feldman.
3065
3066         Inspector: Resources Search Should Search Only Filtered Resources
3067         https://bugs.webkit.org/show_bug.cgi?id=28290
3068
3069         * inspector/front-end/AbstractTimelinePanel.js:
3070         (WebInspector.AbstractTimelinePanel.prototype._updateFilter):
3071         * inspector/front-end/ResourcesPanel.js:
3072         (WebInspector.ResourcesPanel.prototype.get searchableViews):
3073         * inspector/front-end/inspector.js:
3074         (WebInspector.performSearch):
3075         (WebInspector.doPerformSearch):
3076
3077 2010-07-20  Alexander Pavlov  <apavlov@chromium.org>
3078
3079         Reviewed by Pavel Feldman.
3080
3081         Web Inspector: Incorrect absolute URLs in tooltips of links in the ElementsTreeOutline
3082         https://bugs.webkit.org/show_bug.cgi?id=42626
3083
3084         * inspector/front-end/ElementsTreeOutline.js:
3085         (WebInspector.ElementsTreeElement.prototype._attributeHTML):
3086
3087 2010-07-20  Shinichiro Hamaji  <hamaji@chromium.org>
3088
3089         Unreviewed comment fix. The comment in IDL should be //, not # .
3090
3091         Code generator: ensure generated constants match their corresponding enums.
3092         https://bugs.webkit.org/show_bug.cgi?id=42250
3093
3094         * dom/OverflowEvent.idl:
3095
3096 2010-07-20  Hans Wennborg  <hans@chromium.org>
3097
3098         Reviewed by Steve Block.
3099
3100         Add WebCore/bindings/generic/RuntimeEnabledFeatures.cpp to build files
3101         https://bugs.webkit.org/show_bug.cgi?id=42380
3102
3103         RuntimeEnabledFeatures.cpp and .h were moved from bindings/v8 to
3104         bindings/generic a while a go (in r54593), but need to
3105         be added to the build in order to be used.
3106
3107         No new functionality so no new tests.
3108
3109         * Android.mk:
3110         * CMakeLists.txt:
3111         * GNUmakefile.am:
3112         * WebCore.pro:
3113         * WebCore.vcproj/WebCore.vcproj:
3114         * WebCore.xcodeproj/project.pbxproj:
3115         * websockets/WebSocket.cpp: Remove #if USE(V8), as RuntimeEnabledFeatures.cpp is no longer V8-only.
3116         * websockets/WebSocket.h: Ditto.
3117
3118 2010-07-20  Tony Chang  <tony@chromium.org>
3119
3120         Reviewed by Kent Tamura.
3121
3122         clean up style in ClipboardWin and PasteboardWin
3123         https://bugs.webkit.org/show_bug.cgi?id=42609
3124
3125         No new tests since this is just a style cleanup.
3126
3127         * platform/win/ClipboardUtilitiesWin.cpp:
3128         (WebCore::createGlobalData):
3129         (WebCore::markupToCFHTML):
3130         (WebCore::getURL):
3131         (WebCore::getPlainText):
3132         (WebCore::getTextHTML):
3133         (WebCore::fragmentFromFilenames):
3134         (WebCore::containsFilenames):
3135         (WebCore::fragmentFromCFHTML):
3136         (WebCore::fragmentFromHTML):
3137         * platform/win/ClipboardUtilitiesWin.h:
3138         * platform/win/ClipboardWin.cpp:
3139         (WebCore::pathRemoveBadFSCharacters):
3140         (WebCore::createGlobalHDropContent):
3141         (WebCore::createGlobalUrlFileDescriptor):
3142         (WebCore::writeURL):
3143         (WebCore::ClipboardWin::clearData):
3144         (WebCore::ClipboardWin::clearAllData):
3145         (WebCore::ClipboardWin::getData):
3146         (WebCore::ClipboardWin::types):
3147         (WebCore::ClipboardWin::declareAndWriteDragImage):
3148         (WebCore::ClipboardWin::writeRange):
3149         * platform/win/ClipboardWin.h:
3150         (WebCore::ClipboardWin::create):
3151         (WebCore::ClipboardWin::dataObject):
3152         * platform/win/PasteboardWin.cpp:
3153         (WebCore::PasteboardOwnerWndProc):
3154         (WebCore::Pasteboard::writeSelection):
3155         (WebCore::Pasteboard::writeURL):
3156         (WebCore::Pasteboard::plainText):
3157         (WebCore::Pasteboard::documentFragment):
3158
3159 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
3160
3161         Reviewed by Pavel Feldman.
3162
3163         Web Inspector: jump to source is broken for call frames logged by console.trace
3164         https://bugs.webkit.org/show_bug.cgi?id=42619
3165
3166         * inspector/front-end/ConsoleView.js:
3167         (WebInspector.ConsoleMessage.prototype._populateStackTraceTreeElement):
3168         * inspector/front-end/treeoutline.js: allow passing HTML element as a title.
3169         (TreeElement.prototype.set title):
3170         (TreeElement.prototype.set shouldRefreshChildren):
3171         (TreeElement.prototype._setListItemNodeContent):
3172         (TreeElement.prototype._attach):
3173
3174 2010-07-20  Tony Chang  <tony@chromium.org>
3175
3176         Reviewed by Dan Bernstein.
3177
3178         [chromium] crash in Position::getInlineBoxAndOffset
3179         https://bugs.webkit.org/show_bug.cgi?id=42202
3180
3181         Test: editing/selection/firstRect-crash.html
3182         Manual test: win/contextmenu-key2.html
3183
3184         * page/EventHandler.cpp: Code no longer needed
3185         * page/Frame.cpp:
3186         (WebCore::Frame::firstRectForRange): Normalize Positions to VisiblePositions because
3187             Positions may be pointing to nodes that have no renderer.  If there's no renderer,
3188             getInlineBoxAndOffset will crash.
3189
3190 2010-07-20  Adam Barth  <abarth@webkit.org>
3191
3192         Reviewed by Darin Adler.
3193
3194         Foster parenting depends on the current element at insertion time
3195         https://bugs.webkit.org/show_bug.cgi?id=42599
3196
3197         We need to consider the current element at insertion-time when deciding
3198         whether to redirect insertion to the foster parent.  Previously, we
3199         were considering the current element *both* at insertion-time and at
3200         the time we created the guard.
3201
3202         * html/HTMLConstructionSite.h:
3203         (WebCore::HTMLConstructionSite::RedirectToFosterParentGuard::RedirectToFosterParentGuard):
3204         * html/HTMLTreeBuilder.cpp:
3205         (WebCore::HTMLTreeBuilder::processStartTagForInTable):
3206         (WebCore::HTMLTreeBuilder::processEndTagForInTable):
3207         (WebCore::HTMLTreeBuilder::defaultForInTableText):
3208
3209 2010-07-20  Matthew Delaney  <mdelaney@apple.com>
3210
3211         Reviewed by Kenneth Rohde Christiansen.
3212
3213         Failing 2d.path.stroke.prune.curve philip canvas test
3214         https://bugs.webkit.org/show_bug.cgi?id=42190
3215
3216         * html/canvas/CanvasRenderingContext2D.cpp:
3217         (WebCore::CanvasRenderingContext2D::closePath): Added check to make sure there's a non-trivial path to close. Since there is currently no way to check if the current point is the start point, or similarly if there is only 1 point in the current subpath (since these are both sufficient conditions for a trivial subpath), then checking that the bounding rectangle has both zero width and height proves also to be a sufficient condition for a trivial path.
3218         (WebCore::CanvasRenderingContext2D::quadraticCurveTo): Added in simple bounds as per the spec.
3219         (WebCore::CanvasRenderingContext2D::bezierCurveTo): Added in simple bounds as per the spec.
3220         * platform/graphics/cg/PathCG.cpp:
3221         (WebCore::Path::closeSubpath): Moved the check for an empty path up on level to make it platform independent and remove redundancy.
3222
3223 2010-07-19  Victoria Kirst  <vrk@google.com>
3224
3225         Reviewed by David Levin.
3226
3227         Added a simple implementation of VideoLayerChromium. Uses the
3228         LayerChromium::updateTextureRect() to send video frames to the
3229         GPU.
3230         https://bugs.webkit.org/show_bug.cgi?id=42234
3231
3232         * WebCore.gypi: Added include for VideoLayerChromium.
3233         * platform/graphics/chromium/GraphicsLayerChromium.cpp:
3234         (WebCore::GraphicsLayerChromium::setContentsToMedia): Implemented
3235         setContentsToMedia, though it does not seem to trigger a repaint
3236         correctly.
3237         * platform/graphics/chromium/GraphicsLayerChromium.h:
3238         * platform/graphics/chromium/VideoLayerChromium.cpp: Added.
3239         (WebCore::VideoLayerChromium::create):
3240         (WebCore::VideoLayerChromium::VideoLayerChromium):
3241         (WebCore::VideoLayerChromium::updateTextureContents):
3242         * platform/graphics/chromium/VideoLayerChromium.h: Added.
3243         (WebCore::VideoLayerChromium::drawsContent):
3244
3245 2010-07-19  Dirk Schulze  <krit@webkit.org>
3246
3247         Reviewed by Nikolas Zimmermann.
3248
3249         SVG CleanUp of SVGPathData parsing
3250         https://bugs.webkit.org/show_bug.cgi?id=41410
3251
3252         Cleanup the parsing code for SVG Paths. Move classes out of the convoluted SVGParserUtilities.cpp
3253         in their own files. Make use of WebCore specific objects in parsing code like FloatPoint, AffineTransform,
3254         etc. instead of using home-brewn solutions.
3255         The SVGPathParser parses a path data string and delivers the parsed segments and values to the SVGPathConsumer.
3256         SVGPathConsumer is the base class for SVGPathBuilder and SVGPathSegListBuilder, that either build the platform
3257         Path object or a SVGPathSegList out of the segments.
3258         We're now directly parsing floats instead of truncating precision to float afterwards.
3259
3260         SVG Path with an arc with radius of 0 does not render
3261         https://bugs.webkit.org/show_bug.cgi?id=40448
3262
3263         If one of the radii on the elliptic arc are zero, we should draw a line from the starting point to
3264         the end point according to the spec. Fixed this bug with this patch, because an is zero check and 
3265         an DRT check was neccessary with the current clean-up.
3266         Extended svg/dom/path-parser.xhml to check the correct behavior.
3267
3268         * Android.mk:
3269         * GNUmakefile.am:
3270         * WebCore.gypi:
3271         * WebCore.pro:
3272         * WebCore.vcproj/WebCore.vcproj:
3273         * WebCore.xcodeproj/project.pbxproj:
3274         * platform/graphics/FloatPoint.h:
3275         (WebCore::FloatPoint::move): Follow WebKit style and use multiple lines for the function.
3276         (WebCore::FloatPoint::scale): Scale FloatPoint.
3277         (WebCore::operator+=):
3278         (WebCore::operator+): Add two FloatPoints and give back the sum as FloatPoint.
3279         * svg/SVGAllInOne.cpp: Added new created files.
3280         * svg/SVGAnimateElement.cpp:
3281         (WebCore::SVGAnimateElement::calculateFromAndToValues): Use new PathParser to create a PathSegList.
3282         * svg/SVGAnimateMotionElement.cpp:
3283         (WebCore::SVGAnimateMotionElement::parseMappedAttribute): Use new PathParser to create a Path.
3284         * svg/SVGGlyphElement.cpp:
3285         (WebCore::parsePathData): Use new PathParser to create a Path.
3286         * svg/SVGParserUtilities.cpp:
3287         (WebCore::parseArcFlag): Removed parseArcFlag(double&), we parse in float now.
3288         * svg/SVGParserUtilities.h: Removed path parsing code and a lot of unneccessary includes.
3289         * svg/SVGPathBuilder.cpp: Added.
3290         (WebCore::SVGPathBuilder::SVGPathBuilder):
3291         (WebCore::SVGPathBuilder::build): Builds a normalized Path.
3292         (WebCore::SVGPathBuilder::moveTo):
3293         (WebCore::SVGPathBuilder::lineTo):
3294         (WebCore::SVGPathBuilder::curveToCubic):
3295         (WebCore::SVGPathBuilder::closePath):
3296         * svg/SVGPathBuilder.h: Added. We just create normalized Paths, so this functions should never be reached.
3297         (WebCore::SVGPathBuilder::lineToHorizontal):
3298         (WebCore::SVGPathBuilder::lineToVertical):
3299         (WebCore::SVGPathBuilder::curveToCubicSmooth):
3300         (WebCore::SVGPathBuilder::curveToQuadratic):
3301         (WebCore::SVGPathBuilder::curveToQuadraticSmooth):
3302         (WebCore::SVGPathBuilder::arcTo):
3303         * svg/SVGPathConsumer.h: Added. Base class of SVGPathBuilder and SVGPathSegListBuilder.
3304         (WebCore::):
3305         (WebCore::SVGPathConsumer::SVGPathConsumer):
3306         (WebCore::SVGPathConsumer::~SVGPathConsumer):
3307         * svg/SVGPathElement.cpp:
3308         (WebCore::SVGPathElement::parseMappedAttribute): Use new PathParser to create a PathSegList.
3309         * svg/SVGPathParser.cpp: Added.
3310         (WebCore::SVGPathParser::SVGPathParser):
3311         (WebCore::SVGPathParser::~SVGPathParser):
3312         (WebCore::SVGPathParser::parseClosePathSegment):
3313         (WebCore::SVGPathParser::parseMoveToSegment):
3314         (WebCore::SVGPathParser::parseLineToSegment):
3315         (WebCore::SVGPathParser::parseLineToHorizontalSegment):
3316         (WebCore::SVGPathParser::parseLineToVerticalSegment):
3317         (WebCore::SVGPathParser::parseCurveToCubicSegment):
3318         (WebCore::SVGPathParser::parseCurveToCubicSmoothSegment):
3319         (WebCore::SVGPathParser::parseCurveToQuadraticSegment):
3320         (WebCore::SVGPathParser::parseCurveToQuadraticSmoothSegment):
3321         (WebCore::SVGPathParser::parseArcToSegment):
3322         (WebCore::SVGPathParser::parsePathDataString):
3323         (WebCore::SVGPathParser::decomposeArcToCubic): Normalizes an arc to multiple cubic curves.
3324         * svg/SVGPathParser.h: Added.
3325         * svg/SVGPathSegListBuilder.cpp: Added.
3326         (WebCore::SVGPathSegListBuilder::SVGPathSegListBuilder):
3327         (WebCore::SVGPathSegListBuilder::build):
3328         (WebCore::SVGPathSegListBuilder::moveTo):
3329         (WebCore::SVGPathSegListBuilder::lineTo):
3330         (WebCore::SVGPathSegListBuilder::lineToHorizontal):
3331         (WebCore::SVGPathSegListBuilder::lineToVertical):
3332         (WebCore::SVGPathSegListBuilder::curveToCubic):
3333         (WebCore::SVGPathSegListBuilder::curveToCubicSmooth):
3334         (WebCore::SVGPathSegListBuilder::curveToQuadratic):
3335         (WebCore::SVGPathSegListBuilder::curveToQuadraticSmooth):
3336         (WebCore::SVGPathSegListBuilder::arcTo):
3337         (WebCore::SVGPathSegListBuilder::closePath):
3338         * svg/SVGPathSegListBuilder.h: Added.
3339
3340 2010-07-15  Yuzo Fujishima  <yuzo@google.com>
3341
3342         Reviewed by Dan Bernstein.
3343
3344         Fix for Bug 42342 - Font download error for an @font-face rule invalidates other @font-face rules for the same font-family
3345         https://bugs.webkit.org/show_bug.cgi?id=42342
3346
3347         Test: fast/css/font-face-download-error.html
3348
3349         * css/CSSSegmentedFontFace.cpp:
3350         (WebCore::CSSSegmentedFontFace::isValid): Valid if at least one font
3351         face is valid.
3352         (WebCore::CSSSegmentedFontFace::getFontData): Check validity for each
3353         font face.
3354         * css/CSSSegmentedFontFace.h: Make isValid private.
3355
3356 2010-07-19  Kent Tamura  <tkent@chromium.org>
3357
3358         Reviewed by Shinichiro Hamaji.
3359
3360         [Chromium] Fix style errors of RenderThemeChromiumWin.cpp
3361         https://bugs.webkit.org/show_bug.cgi?id=42568
3362
3363         * rendering/RenderThemeChromiumWin.cpp:
3364         (WebCore::):
3365         (WebCore::getNonClientMetrics):
3366         (WebCore::systemFontSize):
3367         (WebCore::pointsToPixels):
3368         (WebCore::querySystemBlinkInterval):
3369         (WebCore::RenderThemeChromiumWin::platformActiveSelectionBackgroundColor):
3370         (WebCore::RenderThemeChromiumWin::platformInactiveSelectionBackgroundColor):
3371         (WebCore::RenderThemeChromiumWin::platformActiveSelectionForegroundColor):
3372         (WebCore::RenderThemeChromiumWin::platformActiveTextSearchHighlightColor):
3373         (WebCore::RenderThemeChromiumWin::paintButton):
3374         (WebCore::RenderThemeChromiumWin::paintSliderTrack):
3375         (WebCore::RenderThemeChromiumWin::paintMenuList):
3376         (WebCore::RenderThemeChromiumWin::paintTextFieldInternal):
3377         (WebCore::RenderThemeChromiumWin::paintInnerSpinButton):
3378
3379 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3380
3381         (Hopefully the last) Windows build fix.
3382         
3383         Version of CoreServices.h in WebKit Support Libraries uses
3384         pragma once, unlike the version I have locally. Switch based
3385         on __COLORSYNCDEPRECATED__ instead.
3386
3387         * WebCorePrefix.h:
3388
3389 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3390
3391         Windows build fix.
3392
3393         * WebCorePrefix.h:
3394
3395 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3396
3397         Build fix.
3398
3399         * WebCorePrefix.h:
3400
3401 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3402
3403         Build fix.
3404
3405         * WebCorePrefix.h:
3406
3407 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3408
3409         Windows Build fixes for new ColorSync API.
3410         We support both new and old APIs, since the newer headers aren't in the tree yet.
3411
3412         * WebCorePrefix.h: Removed include of CoreServices.h. Included via ColorSyncPriv.h instead, since header may not be present.
3413         * platform/graphics/cg/ColorCG.cpp:
3414         (WebCore::createCGColor): Conditionally use new ColorSync API.
3415         * platform/graphics/opentype/OpenTypeUtilities.cpp: Define Fixed if CoreServices.h doesn't.
3416         * platform/graphics/win/GraphicsLayerCACF.cpp:
3417         (WebCore::GraphicsLayerCACF::updateLayerDrawsContent): Use 0 instead of nil, since nil is no longer present via CoreServices.h.
3418         * platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp: Include AssertMacros.h.
3419         * platform/network/cf/ResourceErrorCF.cpp:
3420         (WebCore::ResourceError::operator CFErrorRef): Use 0 instead of nil, since nil is no longer present via CoreServices.h.
3421
3422 2010-07-19  Tony Gentilcore  <tonyg@chromium.org>
3423
3424         Reviewed by Dimitri Glazkov.
3425
3426         Only set unloadEventEnd when the unload event is actually fired
3427         https://bugs.webkit.org/show_bug.cgi?id=42607
3428
3429         r63689 introduced this ASSERT and it began failing on Qt and Mac debug.
3430         For some reason, it did not fail on Windows (which is where I was
3431         testing). The ASSERT was disabled in r63699.
3432
3433         Test: page-cache related layout tests don't crash in debug mode on Mac.
3434
3435         * loader/FrameLoader.cpp:
3436         (WebCore::FrameLoader::stopLoading):
3437
3438 2010-07-19  Andreas Kling  <andreas.kling@nokia.com>
3439
3440         Reviewed by Kenneth Rohde Christiansen.
3441
3442         [Qt] Don't unnecessarily copy QPainterPath in fillPath() and strokePath()
3443         https://bugs.webkit.org/show_bug.cgi?id=42513
3444
3445         Avoid making unnecessary deep-copies of QPainterPaths that will
3446         be discarded after use.
3447
3448         * platform/graphics/qt/GraphicsContextQt.cpp:
3449         (WebCore::GraphicsContext::fillPath):
3450         (WebCore::GraphicsContext::strokePath):
3451
3452 2010-07-19  Kenneth Russell  <kbr@google.com>
3453
3454         Reviewed by Nate Chapin.
3455
3456         WebGL rendering results must be made available to Canvas.toDataURL and 2D drawImage
3457         https://bugs.webkit.org/show_bug.cgi?id=34719
3458
3459         Fixed compiler warning introduced by original patch. No new tests;
3460         covered by existing tests.
3461
3462         * platform/graphics/mac/GraphicsContext3DMac.mm:
3463         (WebCore::GraphicsContext3D::paintRenderingResultsToCanvas):
3464
3465 2010-07-19  Anders Carlsson  <andersca@apple.com>
3466
3467         Reviewed by Darin Adler, Adam Roben, Dan Bernstein and Sam Weinig.
3468
3469         Handle NP_ASFILE and NP_ASFILEONLY transfer modes
3470         https://bugs.webkit.org/show_bug.cgi?id=42587
3471
3472         * WebCore.exp.in:
3473         Export functions from FileSystem.h
3474         
3475         * platform/mac/FileSystemMac.mm:
3476         (WebCore::openTemporaryFile):
3477         Try to create a temporary file using mkstemp.
3478
3479 2010-07-19  Anders Carlsson  <andersca@apple.com>
3480
3481         Reviewed by Sam Weinig.
3482
3483         Implement NPN_PostURLNotify
3484         https://bugs.webkit.org/show_bug.cgi?id=42602
3485
3486         Export symbols needed by WebKit2.
3487
3488         * WebCore.exp.in:
3489
3490 2010-07-19  Tony Gentilcore  <tonyg@chromium.org>
3491
3492         Unreviewed build fix.
3493
3494         Remove overzealous ASSERT from r63689
3495         https://bugs.webkit.org/show_bug.cgi?id=42606
3496
3497         No new tests because no new functionality.
3498
3499         * loader/FrameLoader.cpp:
3500         (WebCore::FrameLoader::stopLoading):
3501
3502 2010-07-19  Joseph Pecoraro  <joepeck@webkit.org>
3503
3504         Reviewed by Mark Rowe.
3505
3506         Web Inspector: Do Not Copy *.re2js Inspector Resources in XCode Build Phase
3507         https://bugs.webkit.org/show_bug.cgi?id=42601
3508
3509         Remove *.re2js files after copying them over in the Build Phase. We do
3510         the same to remove the WebKit.qrc file.
3511
3512         * WebCore.xcodeproj/project.pbxproj:
3513
3514 2010-07-17  Tony Gentilcore  <tonyg@chromium.org>
3515
3516         Reviewed by Darin Fisher.
3517
3518         [Web Timing] Move times to DocumentLoader and fix bugs in mark points
3519         https://bugs.webkit.org/show_bug.cgi?id=42512
3520
3521         Test: fast/dom/webtiming-navigate-within-document.html
3522
3523         * loader/DocumentLoader.h: Move the FrameLoadTimeline (now call DocumentLoadTiming) to the DocumentLoader.
3524         (WebCore::DocumentLoader::documentLoadTiming):
3525         * loader/FrameLoader.cpp:
3526         (WebCore::FrameLoader::stopLoading): Set unloadEventEnd on the provisional DocumentLoader. Add some ASSERTs to tighten things up.
3527         (WebCore::FrameLoader::loadWithDocumentLoader): This was not the right place to set navigationStart. Setting it here caused it to be set before the unload form prompt and caused it to be reset when navigating within the document.
3528         (WebCore::FrameLoader::finishedLoading): Set responseEnd on the active DocumentLoader.
3529         (WebCore::FrameLoader::continueLoadAfterWillSubmitForm): This is the right place for navigationStart as defined by the spec.
3530         * loader/FrameLoader.h: Get rid of FrameLoadTimeline.
3531         * loader/FrameLoaderTypes.h: Rename FrameLoadTimeline to DocumentLoadTiming. It is even more apparent this doesn't belong in this file now. I am planning to submit a patch moving it out ASAP, but didn't want to muddy this patch with all those build files.
3532         (WebCore::DocumentLoadTiming::DocumentLoadTiming):
3533         * loader/MainResourceLoader.cpp:
3534         (WebCore::MainResourceLoader::willSendRequest): Move fetchStart out of this method to load(), and rewrite setting of redirectStart, redirectEnd, and redirectCount to be more readable.
3535         (WebCore::MainResourceLoader::load): Set fetchStart slightly earlier here and tighten it up with some ASSERTs.
3536         * page/DOMWindow.cpp:
3537         (WebCore::DOMWindow::dispatchLoadEvent): Set loadEventStart and loadEventEnd on the DocumentLoader.
3538         * page/Navigation.cpp:
3539         (WebCore::Navigation::redirectCount): Retrieve redirectCount from the DocumentLoader.
3540         * page/Timing.cpp:
3541         (WebCore::getPossiblySkewedTimeInKnownRange): The skew problem turned out to be due to the fact that chromium's currentTime() implementation only syncs to the system time every 60 seconds. So absolute times across threads may be skewed slightly. I resolved this temporarily by clipping the time from another thread into a known bound. A better long term solution is probably to add a currentTimeFromSystemTime() method and call that for web timing marks.
3542         (WebCore::Timing::navigationStart):
3543         (WebCore::Timing::unloadEventEnd):
3544         (WebCore::Timing::redirectStart):
3545         (WebCore::Timing::redirectEnd):
3546         (WebCore::Timing::fetchStart):
3547         (WebCore::Timing::domainLookupStart):
3548         (WebCore::Timing::domainLookupEnd):
3549         (WebCore::Timing::connectStart):
3550         (WebCore::Timing::connectEnd):
3551         (WebCore::Timing::requestStart):
3552         (WebCore::Timing::requestEnd):
3553         (WebCore::Timing::responseStart):
3554         (WebCore::Timing::responseEnd):
3555         (WebCore::Timing::loadEventStart):
3556         (WebCore::Timing::loadEventEnd):
3557         (WebCore::Timing::documentLoader):
3558         (WebCore::Timing::documentLoadTiming):
3559         (WebCore::Timing::resourceLoadTiming):
3560         (WebCore::Timing::resourceLoadTimeRelativeToAbsolute): Ensure requestTime is in the range of fetchStart to responseEnd.
3561         * page/Timing.h:
3562
3563 2010-07-19  Chris Marrin  <cmarrin@apple.com>
3564
3565         Reviewed by Darin Adler.
3566
3567         https://bugs.webkit.org/show_bug.cgi?id=42118
3568         Disable WebGL on Leopard for now. 
3569
3570         LayoutTests fail on some graphics hardware on Leopard because one of the features we use,
3571         GL_ARB_framebuffer_object, is not universally available in Leopard like it is in
3572         SnowLeopard. This will allow LayoutTests to pass on Leopard until we add logic to use a
3573         software OpenGL driver on machines without this support.
3574
3575         * Configurations/FeatureDefines.xcconfig:
3576
3577 2010-07-19  Eric Carlson  <eric.carlson@apple.com>
3578
3579         Reviewed by Sam Weinig.
3580
3581         Remove HTML5 media element 'load' event
3582         https://bugs.webkit.org/show_bug.cgi?id=30464
3583         <rdar://problem/5650561>
3584
3585         * html/HTMLMediaElement.cpp:
3586         (WebCore::HTMLMediaElement::HTMLMediaElement): Initialize m_completelyLoaded.
3587         (WebCore::HTMLMediaElement::parseMappedAttribute): Don't deal with 'load' event.
3588         (WebCore::HTMLMediaElement::prepareForLoad): Set m_completelyLoaded to false.
3589         (WebCore::HTMLMediaElement::setNetworkState): Don't post 'load' event.
3590         (WebCore::HTMLMediaElement::progressEventTimerFired): Bail if m_networkState != NETWORK_LOADING.
3591         (WebCore::HTMLMediaElement::userCancelledLoad): No more NETWORK_LOADED state.
3592         * html/HTMLMediaElement.h:
3593         (WebCore::HTMLMediaElement::):
3594         * html/HTMLMediaElement.idl: Remove NETWORK_LOADING.
3595
3596 2010-07-19  Alexey Proskuryakov  <ap@apple.com>
3597
3598         Reviewed by Darin Adler.
3599
3600         https://bugs.webkit.org/show_bug.cgi?id=40996
3601         Progress event should not be fired during synchronous XMLHttpRequest
3602
3603         https://bugs.webkit.org/show_bug.cgi?id=17502
3604         Assertion failure when trying to restart a sync XMLHttpRequest as an async one from onreadystatechange
3605
3606         Tests: http/tests/xmlhttprequest/xmlhttprequest-sync-no-progress-events.html
3607                http/tests/xmlhttprequest/xmlhttprequest-sync-vs-async-assertion-failure.html
3608
3609         * xml/XMLHttpRequest.cpp:
3610         (WebCore::XMLHttpRequest::callReadyStateChangeListener): We now only dispatch readystatechange
3611         event for synchronous requests in states UNSENT, OPENED and DONE. I'm not sure what exactly
3612         the spec draft says about readystatechange for sync requests, but this seems to be the most
3613         logical and backwards compatible behavior.
3614         (WebCore::XMLHttpRequest::didReceiveData): Don't dispatch progress events for sync requests.
3615         Note that we already don't dispatch upload progress events for those.
3616
3617 2010-07-19  Dan Bernstein  <mitz@apple.com>
3618
3619         Reviewed by Simon Fraser.
3620
3621         <rdar://problem/7232109> Unpainted white area appears at the edge of the page when body has bg color
3622         https://bugs.webkit.org/show_bug.cgi?id=34913
3623
3624         Tests: fast/repaint/view-background-from-body-1.html
3625                fast/repaint/view-background-from-body-2.html
3626
3627         * rendering/RenderBox.cpp:
3628         (WebCore::RenderBox::styleWillChange): If this is the body renderer and its current style is
3629         null, repaint the view, similarly to how the view is repainted for any repaint-or-higher
3630         style changes.
3631         * rendering/RenderObjectChildList.cpp:
3632         (WebCore::RenderObjectChildList::removeChildNode): If the removed child is the body renderer,
3633         repaint the view, in case the body’s background was propagated to the view.
3634
3635 2010-07-19  Adam Roben  <aroben@apple.com>